From c9cbf3d3b35f198fab39e98d696312dd0b97a69a Mon Sep 17 00:00:00 2001 From: Dan Streetman Date: Sat, 10 Jan 2009 23:44:22 -0800 Subject: [PATCH 01/92] Input: usbtouchscreen - allow reporting calibrated data This patch adds a module parameter to report either the raw coordinate data or the hardware-calibrated coordinate data for MicroTouch/3M touchscreens. The default is set to the raw coordinates for backwards compatibilty. Signed-off-by: Dan Streetman Signed-off-by: Dmitry Torokhov --- drivers/input/touchscreen/usbtouchscreen.c | 19 +++++++++++++++++-- 1 file changed, 17 insertions(+), 2 deletions(-) diff --git a/drivers/input/touchscreen/usbtouchscreen.c b/drivers/input/touchscreen/usbtouchscreen.c index 5080b26ba160..6d27a1d661e6 100644 --- a/drivers/input/touchscreen/usbtouchscreen.c +++ b/drivers/input/touchscreen/usbtouchscreen.c @@ -60,6 +60,10 @@ static int swap_xy; module_param(swap_xy, bool, 0644); MODULE_PARM_DESC(swap_xy, "If set X and Y axes are swapped."); +static int hwcalib_xy; +module_param(hwcalib_xy, bool, 0644); +MODULE_PARM_DESC(hwcalib_xy, "If set hw-calibrated X/Y are used if available"); + /* device specifc data/functions */ struct usbtouch_usb; struct usbtouch_device_info { @@ -260,8 +264,13 @@ static int panjit_read_data(struct usbtouch_usb *dev, unsigned char *pkt) static int mtouch_read_data(struct usbtouch_usb *dev, unsigned char *pkt) { - dev->x = (pkt[8] << 8) | pkt[7]; - dev->y = (pkt[10] << 8) | pkt[9]; + if (hwcalib_xy) { + dev->x = (pkt[4] << 8) | pkt[3]; + dev->y = 0xffff - ((pkt[6] << 8) | pkt[5]); + } else { + dev->x = (pkt[8] << 8) | pkt[7]; + dev->y = (pkt[10] << 8) | pkt[9]; + } dev->touch = (pkt[2] & 0x40) ? 1 : 0; return 1; @@ -294,6 +303,12 @@ static int mtouch_init(struct usbtouch_usb *usbtouch) return ret; } + /* Default min/max xy are the raw values, override if using hw-calib */ + if (hwcalib_xy) { + input_set_abs_params(usbtouch->input, ABS_X, 0, 0xffff, 0, 0); + input_set_abs_params(usbtouch->input, ABS_Y, 0, 0xffff, 0, 0); + } + return 0; } #endif From 520abcdeb13de23e22b7d4367b1c3b136ef3b108 Mon Sep 17 00:00:00 2001 From: Daniel Mierswa Date: Sat, 10 Jan 2009 23:44:22 -0800 Subject: [PATCH 02/92] Input: atkbd - make forced_release_keys[] static Signed-off-by: Andrew Morton Signed-off-by: Dmitry Torokhov --- drivers/input/keyboard/atkbd.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/input/keyboard/atkbd.c b/drivers/input/keyboard/atkbd.c index f6e9f39a527b..6df9ba186121 100644 --- a/drivers/input/keyboard/atkbd.c +++ b/drivers/input/keyboard/atkbd.c @@ -839,7 +839,7 @@ static void atkbd_disconnect(struct serio *serio) */ static void atkbd_dell_laptop_keymap_fixup(struct atkbd *atkbd) { - const unsigned int forced_release_keys[] = { + static const unsigned int forced_release_keys[] = { 0x85, 0x86, 0x87, 0x88, 0x89, 0x8a, 0x8b, 0x8f, 0x93, }; int i; @@ -856,7 +856,7 @@ static void atkbd_dell_laptop_keymap_fixup(struct atkbd *atkbd) */ static void atkbd_hp_keymap_fixup(struct atkbd *atkbd) { - const unsigned int forced_release_keys[] = { + static const unsigned int forced_release_keys[] = { 0x94, }; int i; From 3c0340b774916caa4149892504169d290c8a720a Mon Sep 17 00:00:00 2001 From: Jean Delvare Date: Sat, 10 Jan 2009 23:44:22 -0800 Subject: [PATCH 03/92] Input: psmouse - make MOUSE_PS2_LIFEBOOK depend on X86 All Fujitsu-Siemens Lifebook systems are x86-based, so we might as well make MOUSE_PS2_LIFEBOOK depend on X86. This will avoid surprising things like: arch/arm/configs/s3c2410_defconfig:CONFIG_MOUSE_PS2_LIFEBOOK=y Signed-off-by: Jean Delvare Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/Kconfig | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/input/mouse/Kconfig b/drivers/input/mouse/Kconfig index 093c8c1bca74..9705f3a00a3d 100644 --- a/drivers/input/mouse/Kconfig +++ b/drivers/input/mouse/Kconfig @@ -70,7 +70,7 @@ config MOUSE_PS2_SYNAPTICS config MOUSE_PS2_LIFEBOOK bool "Fujitsu Lifebook PS/2 mouse protocol extension" if EMBEDDED default y - depends on MOUSE_PS2 + depends on MOUSE_PS2 && X86 help Say Y here if you have a Fujitsu B-series Lifebook PS/2 TouchScreen connected to your system. From 74ca11c2056d01d9ebb3615cd781a148450c3c82 Mon Sep 17 00:00:00 2001 From: Bastien Nocera Date: Sat, 10 Jan 2009 23:44:22 -0800 Subject: [PATCH 04/92] Input: uvc - the button on the camera is KEY_CAMERA Cameras should generate KEY_CAMERA, not BTN_0. Also call input_sync() on the device once the button has been pressed. Signed-off-by: Andrew Morton Signed-off-by: Dmitry Torokhov --- drivers/media/video/uvc/uvc_status.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/drivers/media/video/uvc/uvc_status.c b/drivers/media/video/uvc/uvc_status.c index 5d60b264d59a..fdf7b4187ee1 100644 --- a/drivers/media/video/uvc/uvc_status.c +++ b/drivers/media/video/uvc/uvc_status.c @@ -47,8 +47,8 @@ static int uvc_input_init(struct uvc_device *dev) usb_to_input_id(udev, &input->id); input->dev.parent = &dev->intf->dev; - set_bit(EV_KEY, input->evbit); - set_bit(BTN_0, input->keybit); + __set_bit(EV_KEY, input->evbit); + __set_bit(KEY_CAMERA, input->keybit); if ((ret = input_register_device(input)) < 0) goto error; @@ -71,8 +71,10 @@ static void uvc_input_cleanup(struct uvc_device *dev) static void uvc_input_report_key(struct uvc_device *dev, unsigned int code, int value) { - if (dev->input) + if (dev->input) { input_report_key(dev->input, code, value); + input_sync(dev->input); + } } #else @@ -97,7 +99,7 @@ static void uvc_event_streaming(struct uvc_device *dev, __u8 *data, int len) return; uvc_trace(UVC_TRACE_STATUS, "Button (intf %u) %s len %d\n", data[1], data[3] ? "pressed" : "released", len); - uvc_input_report_key(dev, BTN_0, data[3]); + uvc_input_report_key(dev, KEY_CAMERA, data[3]); } else { uvc_trace(UVC_TRACE_STATUS, "Stream %u error event %02x %02x " "len %d.\n", data[1], data[2], data[3], len); From f2d8dc75a14479f8803a70cf637b5d79a3bb87f1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Mon, 12 Jan 2009 22:29:23 -0800 Subject: [PATCH 05/92] Input: corgikbd - mark probe function as __devinit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A pointer to corgikbd_probe is passed to the core via platform_driver_register and so the function must not disappear when the .init sections are discarded. Otherwise (if also having HOTPLUG=y) unbinding and binding a device to the driver via sysfs will result in an oops as does a device being registered late. An alternative to this patch is using platform_driver_probe instead of platform_driver_register plus removing the pointer to the probe function from the struct platform_driver. [dtor@mail.ru: fixed some more section markups] Signed-off-by: Uwe Kleine-König Signed-off-by: Dmitry Torokhov --- drivers/input/keyboard/corgikbd.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/input/keyboard/corgikbd.c b/drivers/input/keyboard/corgikbd.c index c8ed065ea0cb..abb04c82c622 100644 --- a/drivers/input/keyboard/corgikbd.c +++ b/drivers/input/keyboard/corgikbd.c @@ -288,7 +288,7 @@ static int corgikbd_resume(struct platform_device *dev) #define corgikbd_resume NULL #endif -static int __init corgikbd_probe(struct platform_device *pdev) +static int __devinit corgikbd_probe(struct platform_device *pdev) { struct corgikbd *corgikbd; struct input_dev *input_dev; @@ -368,7 +368,7 @@ static int __init corgikbd_probe(struct platform_device *pdev) return err; } -static int corgikbd_remove(struct platform_device *pdev) +static int __devexit corgikbd_remove(struct platform_device *pdev) { int i; struct corgikbd *corgikbd = platform_get_drvdata(pdev); @@ -388,7 +388,7 @@ static int corgikbd_remove(struct platform_device *pdev) static struct platform_driver corgikbd_driver = { .probe = corgikbd_probe, - .remove = corgikbd_remove, + .remove = __devexit_p(corgikbd_remove), .suspend = corgikbd_suspend, .resume = corgikbd_resume, .driver = { @@ -397,7 +397,7 @@ static struct platform_driver corgikbd_driver = { }, }; -static int __devinit corgikbd_init(void) +static int __init corgikbd_init(void) { return platform_driver_register(&corgikbd_driver); } From 840207edfa8b5e5b46e0d268bf33efe71fecea20 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Mon, 12 Jan 2009 22:32:17 -0800 Subject: [PATCH 06/92] Input: corgi_ts - mark probe function as __devinit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A pointer to corgits_probe is passed to the core via platform_driver_register and so the function must not disappear when the .init sections are discarded. Otherwise (if also having HOTPLUG=y) unbinding and binding a device to the driver via sysfs will result in an oops as does a device being registered late. An alternative to this patch is using platform_driver_probe instead of platform_driver_register plus removing the pointer to the probe function from the struct platform_driver. [dtor@mail.ru: fixed some more section markups] Signed-off-by: Uwe Kleine-König Signed-off-by: Dmitry Torokhov --- drivers/input/touchscreen/corgi_ts.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/input/touchscreen/corgi_ts.c b/drivers/input/touchscreen/corgi_ts.c index 65202c9f63ff..3fb51b54fe61 100644 --- a/drivers/input/touchscreen/corgi_ts.c +++ b/drivers/input/touchscreen/corgi_ts.c @@ -268,7 +268,7 @@ static int corgits_resume(struct platform_device *dev) #define corgits_resume NULL #endif -static int __init corgits_probe(struct platform_device *pdev) +static int __devinit corgits_probe(struct platform_device *pdev) { struct corgi_ts *corgi_ts; struct input_dev *input_dev; @@ -343,7 +343,7 @@ static int __init corgits_probe(struct platform_device *pdev) return err; } -static int corgits_remove(struct platform_device *pdev) +static int __devexit corgits_remove(struct platform_device *pdev) { struct corgi_ts *corgi_ts = platform_get_drvdata(pdev); @@ -352,12 +352,13 @@ static int corgits_remove(struct platform_device *pdev) corgi_ts->machinfo->put_hsync(); input_unregister_device(corgi_ts->input); kfree(corgi_ts); + return 0; } static struct platform_driver corgits_driver = { .probe = corgits_probe, - .remove = corgits_remove, + .remove = __devexit_p(corgits_remove), .suspend = corgits_suspend, .resume = corgits_resume, .driver = { @@ -366,7 +367,7 @@ static struct platform_driver corgits_driver = { }, }; -static int __devinit corgits_init(void) +static int __init corgits_init(void) { return platform_driver_register(&corgits_driver); } From d63dab00e85641515a0a060e6de93e6b7f450665 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Tue, 13 Jan 2009 18:20:20 -0800 Subject: [PATCH 07/92] Input: omap-keypad - mark probe function as __devinit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A pointer to omap_kp_probe is passed to the core via platform_driver_register and so the function must not disappear when the .init sections are discarded. Otherwise (if also having HOTPLUG=y) unbinding and binding a device to the driver via sysfs will result in an oops as does a device being registered late. An alternative to this patch is using platform_driver_probe instead of platform_driver_register plus removing the pointer to the probe function from the struct platform_driver. [dtor@mail.ru: fixed some more section markups] Signed-off-by: Uwe Kleine-König Signed-off-by: Dmitry Torokhov --- drivers/input/keyboard/omap-keypad.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/input/keyboard/omap-keypad.c b/drivers/input/keyboard/omap-keypad.c index ec0ebee46069..23b4fd18f851 100644 --- a/drivers/input/keyboard/omap-keypad.c +++ b/drivers/input/keyboard/omap-keypad.c @@ -279,7 +279,7 @@ static int omap_kp_resume(struct platform_device *dev) #define omap_kp_resume NULL #endif -static int __init omap_kp_probe(struct platform_device *pdev) +static int __devinit omap_kp_probe(struct platform_device *pdev) { struct omap_kp *omap_kp; struct input_dev *input_dev; @@ -422,7 +422,7 @@ err1: return -EINVAL; } -static int omap_kp_remove(struct platform_device *pdev) +static int __devexit omap_kp_remove(struct platform_device *pdev) { struct omap_kp *omap_kp = platform_get_drvdata(pdev); @@ -454,7 +454,7 @@ static int omap_kp_remove(struct platform_device *pdev) static struct platform_driver omap_kp_driver = { .probe = omap_kp_probe, - .remove = omap_kp_remove, + .remove = __devexit_p(omap_kp_remove), .suspend = omap_kp_suspend, .resume = omap_kp_resume, .driver = { @@ -463,7 +463,7 @@ static struct platform_driver omap_kp_driver = { }, }; -static int __devinit omap_kp_init(void) +static int __init omap_kp_init(void) { printk(KERN_INFO "OMAP Keypad Driver\n"); return platform_driver_register(&omap_kp_driver); From 27f23336bcc1b08512751c878b3e05ed6b815e86 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Uwe=20Kleine-K=C3=B6nig?= Date: Tue, 13 Jan 2009 18:20:31 -0800 Subject: [PATCH 08/92] Input: spitzkbd - mark probe function as __devinit MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit A pointer to spitzkbd_probe is passed to the core via platform_driver_register and so the function must not disappear when the .init sections are discarded. Otherwise (if also having HOTPLUG=y) unbinding and binding a device to the driver via sysfs will result in an oops as does a device being registered late. An alternative to this patch is using platform_driver_probe instead of platform_driver_register plus removing the pointer to the probe function from the struct platform_driver. [dtor@mail.ru: fixed some more section markups] Signed-off-by: Uwe Kleine-König Signed-off-by: Dmitry Torokhov --- drivers/input/keyboard/spitzkbd.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/drivers/input/keyboard/spitzkbd.c b/drivers/input/keyboard/spitzkbd.c index c48b76a46a58..9d1781a618e9 100644 --- a/drivers/input/keyboard/spitzkbd.c +++ b/drivers/input/keyboard/spitzkbd.c @@ -343,7 +343,7 @@ static int spitzkbd_resume(struct platform_device *dev) #define spitzkbd_resume NULL #endif -static int __init spitzkbd_probe(struct platform_device *dev) +static int __devinit spitzkbd_probe(struct platform_device *dev) { struct spitzkbd *spitzkbd; struct input_dev *input_dev; @@ -444,7 +444,7 @@ static int __init spitzkbd_probe(struct platform_device *dev) return err; } -static int spitzkbd_remove(struct platform_device *dev) +static int __devexit spitzkbd_remove(struct platform_device *dev) { int i; struct spitzkbd *spitzkbd = platform_get_drvdata(dev); @@ -470,7 +470,7 @@ static int spitzkbd_remove(struct platform_device *dev) static struct platform_driver spitzkbd_driver = { .probe = spitzkbd_probe, - .remove = spitzkbd_remove, + .remove = __devexit_p(spitzkbd_remove), .suspend = spitzkbd_suspend, .resume = spitzkbd_resume, .driver = { @@ -479,7 +479,7 @@ static struct platform_driver spitzkbd_driver = { }, }; -static int __devinit spitzkbd_init(void) +static int __init spitzkbd_init(void) { return platform_driver_register(&spitzkbd_driver); } From e8e0c02340f2d3e4f1ea75a136883d8797290929 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Thu, 29 Jan 2009 22:56:08 -0800 Subject: [PATCH 09/92] Input: bf54x-keys - fix debounce time validation Signed-off-by: Roel Kluin Acked-by: Michael Hennerich Signed-off-by: Dmitry Torokhov --- drivers/input/keyboard/bf54x-keys.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/input/keyboard/bf54x-keys.c b/drivers/input/keyboard/bf54x-keys.c index 19284016e0f4..ee855c5202e8 100644 --- a/drivers/input/keyboard/bf54x-keys.c +++ b/drivers/input/keyboard/bf54x-keys.c @@ -209,8 +209,8 @@ static int __devinit bfin_kpad_probe(struct platform_device *pdev) goto out; } - if (!pdata->debounce_time || !pdata->debounce_time > MAX_MULT || - !pdata->coldrive_time || !pdata->coldrive_time > MAX_MULT) { + if (!pdata->debounce_time || pdata->debounce_time > MAX_MULT || + !pdata->coldrive_time || pdata->coldrive_time > MAX_MULT) { printk(KERN_ERR DRV_NAME ": Invalid Debounce/Columdrive Time from pdata\n"); bfin_write_KPAD_MSEL(0xFF0); /* Default MSEL */ From 4e8718a1f960db0c48427f4583f89f4cb62f2480 Mon Sep 17 00:00:00 2001 From: Kay Sievers Date: Thu, 29 Jan 2009 22:56:08 -0800 Subject: [PATCH 10/92] Input: struct device - replace bus_id with dev_name(), dev_set_name() Acked-by: Greg Kroah-Hartman Signed-off-by: Kay Sievers Signed-off-by: Dmitry Torokhov --- drivers/input/serio/ambakmi.c | 4 ++-- drivers/input/serio/gscps2.c | 2 +- drivers/input/serio/sa1111ps2.c | 4 ++-- drivers/input/touchscreen/atmel_tsadcc.c | 2 +- drivers/input/touchscreen/tsc2007.c | 3 ++- 5 files changed, 8 insertions(+), 7 deletions(-) diff --git a/drivers/input/serio/ambakmi.c b/drivers/input/serio/ambakmi.c index b10ffae7c39b..af159c7d60bc 100644 --- a/drivers/input/serio/ambakmi.c +++ b/drivers/input/serio/ambakmi.c @@ -129,8 +129,8 @@ static int amba_kmi_probe(struct amba_device *dev, void *id) io->write = amba_kmi_write; io->open = amba_kmi_open; io->close = amba_kmi_close; - strlcpy(io->name, dev->dev.bus_id, sizeof(io->name)); - strlcpy(io->phys, dev->dev.bus_id, sizeof(io->phys)); + strlcpy(io->name, dev_name(&dev->dev), sizeof(io->name)); + strlcpy(io->phys, dev_name(&dev->dev), sizeof(io->phys)); io->port_data = kmi; io->dev.parent = &dev->dev; diff --git a/drivers/input/serio/gscps2.c b/drivers/input/serio/gscps2.c index adc3bd6e7f7b..bd0f92d9f40f 100644 --- a/drivers/input/serio/gscps2.c +++ b/drivers/input/serio/gscps2.c @@ -359,7 +359,7 @@ static int __init gscps2_probe(struct parisc_device *dev) snprintf(serio->name, sizeof(serio->name), "GSC PS/2 %s", (ps2port->id == GSC_ID_KEYBOARD) ? "keyboard" : "mouse"); - strlcpy(serio->phys, dev->dev.bus_id, sizeof(serio->phys)); + strlcpy(serio->phys, dev_name(&dev->dev), sizeof(serio->phys)); serio->id.type = SERIO_8042; serio->write = gscps2_write; serio->open = gscps2_open; diff --git a/drivers/input/serio/sa1111ps2.c b/drivers/input/serio/sa1111ps2.c index 2ad88780a170..57953c0eb82f 100644 --- a/drivers/input/serio/sa1111ps2.c +++ b/drivers/input/serio/sa1111ps2.c @@ -246,8 +246,8 @@ static int __devinit ps2_probe(struct sa1111_dev *dev) serio->write = ps2_write; serio->open = ps2_open; serio->close = ps2_close; - strlcpy(serio->name, dev->dev.bus_id, sizeof(serio->name)); - strlcpy(serio->phys, dev->dev.bus_id, sizeof(serio->phys)); + strlcpy(serio->name, dev_name(&dev->dev), sizeof(serio->name)); + strlcpy(serio->phys, dev_name(&dev->dev), sizeof(serio->phys)); serio->port_data = ps2if; serio->dev.parent = &dev->dev; ps2if->io = serio; diff --git a/drivers/input/touchscreen/atmel_tsadcc.c b/drivers/input/touchscreen/atmel_tsadcc.c index a89a6a8f05e6..055969e8be13 100644 --- a/drivers/input/touchscreen/atmel_tsadcc.c +++ b/drivers/input/touchscreen/atmel_tsadcc.c @@ -236,7 +236,7 @@ static int __devinit atmel_tsadcc_probe(struct platform_device *pdev) ts_dev->bufferedmeasure = 0; snprintf(ts_dev->phys, sizeof(ts_dev->phys), - "%s/input0", pdev->dev.bus_id); + "%s/input0", dev_name(&pdev->dev)); input_dev->name = "atmel touch screen controller"; input_dev->phys = ts_dev->phys; diff --git a/drivers/input/touchscreen/tsc2007.c b/drivers/input/touchscreen/tsc2007.c index b75dc2990574..4ab070246892 100644 --- a/drivers/input/touchscreen/tsc2007.c +++ b/drivers/input/touchscreen/tsc2007.c @@ -289,7 +289,8 @@ static int tsc2007_probe(struct i2c_client *client, pdata->init_platform_hw(); - snprintf(ts->phys, sizeof(ts->phys), "%s/input0", client->dev.bus_id); + snprintf(ts->phys, sizeof(ts->phys), + "%s/input0", dev_name(&client->dev)); input_dev->name = "TSC2007 Touchscreen"; input_dev->phys = ts->phys; From bc34496d63ec0a669d6825ea42275fd6fcbe9969 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Sun, 1 Feb 2009 16:54:19 -0800 Subject: [PATCH 11/92] Input: pxa930_trkball - fix write timeout handling With a postfix decrement i reaches -1 rather than 0, but after the loop it is tested whether it has become 0. Signed-off-by: Roel Kluin Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/pxa930_trkball.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/input/mouse/pxa930_trkball.c b/drivers/input/mouse/pxa930_trkball.c index a0f45c4fc198..784be69b7112 100644 --- a/drivers/input/mouse/pxa930_trkball.c +++ b/drivers/input/mouse/pxa930_trkball.c @@ -83,7 +83,7 @@ static int write_tbcr(struct pxa930_trkball *trkball, int v) __raw_writel(v, trkball->mmio_base + TBCR); - while (i--) { + while (--i) { if (__raw_readl(trkball->mmio_base + TBCR) == v) break; msleep(1); From 4ab73761faef832f6d378328f79d21e77c62af3b Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Sun, 1 Feb 2009 16:55:45 -0800 Subject: [PATCH 12/92] Input: ambakmi - fix timeout handling in amba_kmi_write() With a postfix decrement timeleft reaches -1 rather than 0, but after the loop it is tested to have become 0. Signed-off-by: Roel Kluin Signed-off-by: Dmitry Torokhov --- drivers/input/serio/ambakmi.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/input/serio/ambakmi.c b/drivers/input/serio/ambakmi.c index af159c7d60bc..e29cdc13a199 100644 --- a/drivers/input/serio/ambakmi.c +++ b/drivers/input/serio/ambakmi.c @@ -57,7 +57,7 @@ static int amba_kmi_write(struct serio *io, unsigned char val) struct amba_kmi_port *kmi = io->port_data; unsigned int timeleft = 10000; /* timeout in 100ms */ - while ((readb(KMISTAT) & KMISTAT_TXEMPTY) == 0 && timeleft--) + while ((readb(KMISTAT) & KMISTAT_TXEMPTY) == 0 && --timeleft) udelay(10); if (timeleft) From b73a77494292b930642fbf87de3e3196593f7593 Mon Sep 17 00:00:00 2001 From: HighPoint Linux Team Date: Thu, 12 Feb 2009 11:28:31 +0800 Subject: [PATCH 13/92] [SCSI] hptiop: Add new PCI device ID Signed-off-by: HighPoint Linux Team Signed-off-by: James Bottomley --- drivers/scsi/hptiop.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/scsi/hptiop.c b/drivers/scsi/hptiop.c index a48e4990fe12..34be88d7afa5 100644 --- a/drivers/scsi/hptiop.c +++ b/drivers/scsi/hptiop.c @@ -1251,6 +1251,7 @@ static struct pci_device_id hptiop_id_table[] = { { PCI_VDEVICE(TTI, 0x3530), (kernel_ulong_t)&hptiop_itl_ops }, { PCI_VDEVICE(TTI, 0x3560), (kernel_ulong_t)&hptiop_itl_ops }, { PCI_VDEVICE(TTI, 0x4322), (kernel_ulong_t)&hptiop_itl_ops }, + { PCI_VDEVICE(TTI, 0x4321), (kernel_ulong_t)&hptiop_itl_ops }, { PCI_VDEVICE(TTI, 0x4210), (kernel_ulong_t)&hptiop_itl_ops }, { PCI_VDEVICE(TTI, 0x4211), (kernel_ulong_t)&hptiop_itl_ops }, { PCI_VDEVICE(TTI, 0x4310), (kernel_ulong_t)&hptiop_itl_ops }, From 1648b11ea7cec5b95e5a71364ac1f40bfef702d0 Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:38:44 -0800 Subject: [PATCH 14/92] [SCSI] cxgb3i: transmit work-request fixes - resize the work-request credit array to be based on skb's MAX_SKB_FRAGS. - split the skb cb into tx and rx portion - increase the default transmit window to 128K. - stop queueing up the outgoing pdus if transmit window is full. Signed-off-by: Karen Xie Reviewed-by: Mike Christie Signed-off-by: James Bottomley --- drivers/scsi/cxgb3i/cxgb3i_offload.c | 146 +++++++++++++++++++-------- drivers/scsi/cxgb3i/cxgb3i_offload.h | 28 +++-- 2 files changed, 121 insertions(+), 53 deletions(-) diff --git a/drivers/scsi/cxgb3i/cxgb3i_offload.c b/drivers/scsi/cxgb3i/cxgb3i_offload.c index a865f1fefe8b..de3b3b614cca 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_offload.c +++ b/drivers/scsi/cxgb3i/cxgb3i_offload.c @@ -23,19 +23,19 @@ #include "cxgb3i_ddp.h" #ifdef __DEBUG_C3CN_CONN__ -#define c3cn_conn_debug cxgb3i_log_info +#define c3cn_conn_debug cxgb3i_log_debug #else #define c3cn_conn_debug(fmt...) #endif #ifdef __DEBUG_C3CN_TX__ -#define c3cn_tx_debug cxgb3i_log_debug +#define c3cn_tx_debug cxgb3i_log_debug #else #define c3cn_tx_debug(fmt...) #endif #ifdef __DEBUG_C3CN_RX__ -#define c3cn_rx_debug cxgb3i_log_debug +#define c3cn_rx_debug cxgb3i_log_debug #else #define c3cn_rx_debug(fmt...) #endif @@ -47,9 +47,9 @@ static int cxgb3_rcv_win = 256 * 1024; module_param(cxgb3_rcv_win, int, 0644); MODULE_PARM_DESC(cxgb3_rcv_win, "TCP receive window in bytes (default=256KB)"); -static int cxgb3_snd_win = 64 * 1024; +static int cxgb3_snd_win = 128 * 1024; module_param(cxgb3_snd_win, int, 0644); -MODULE_PARM_DESC(cxgb3_snd_win, "TCP send window in bytes (default=64KB)"); +MODULE_PARM_DESC(cxgb3_snd_win, "TCP send window in bytes (default=128KB)"); static int cxgb3_rx_credit_thres = 10 * 1024; module_param(cxgb3_rx_credit_thres, int, 0644); @@ -301,8 +301,8 @@ static void act_open_req_arp_failure(struct t3cdev *dev, struct sk_buff *skb) static void skb_entail(struct s3_conn *c3cn, struct sk_buff *skb, int flags) { - CXGB3_SKB_CB(skb)->seq = c3cn->write_seq; - CXGB3_SKB_CB(skb)->flags = flags; + skb_tcp_seq(skb) = c3cn->write_seq; + skb_flags(skb) = flags; __skb_queue_tail(&c3cn->write_queue, skb); } @@ -457,12 +457,9 @@ static unsigned int wrlen __read_mostly; * The number of WRs needed for an skb depends on the number of fragments * in the skb and whether it has any payload in its main body. This maps the * length of the gather list represented by an skb into the # of necessary WRs. - * - * The max. length of an skb is controlled by the max pdu size which is ~16K. - * Also, assume the min. fragment length is the sector size (512), then add - * extra fragment counts for iscsi bhs and payload padding. + * The extra two fragments are for iscsi bhs and payload padding. */ -#define SKB_WR_LIST_SIZE (16384/512 + 3) +#define SKB_WR_LIST_SIZE (MAX_SKB_FRAGS + 2) static unsigned int skb_wrs[SKB_WR_LIST_SIZE] __read_mostly; static void s3_init_wr_tab(unsigned int wr_len) @@ -485,7 +482,7 @@ static void s3_init_wr_tab(unsigned int wr_len) static inline void reset_wr_list(struct s3_conn *c3cn) { - c3cn->wr_pending_head = NULL; + c3cn->wr_pending_head = c3cn->wr_pending_tail = NULL; } /* @@ -496,7 +493,7 @@ static inline void reset_wr_list(struct s3_conn *c3cn) static inline void enqueue_wr(struct s3_conn *c3cn, struct sk_buff *skb) { - skb_wr_data(skb) = NULL; + skb_tx_wr_next(skb) = NULL; /* * We want to take an extra reference since both us and the driver @@ -509,10 +506,22 @@ static inline void enqueue_wr(struct s3_conn *c3cn, if (!c3cn->wr_pending_head) c3cn->wr_pending_head = skb; else - skb_wr_data(skb) = skb; + skb_tx_wr_next(c3cn->wr_pending_tail) = skb; c3cn->wr_pending_tail = skb; } +static int count_pending_wrs(struct s3_conn *c3cn) +{ + int n = 0; + const struct sk_buff *skb = c3cn->wr_pending_head; + + while (skb) { + n += skb->csum; + skb = skb_tx_wr_next(skb); + } + return n; +} + static inline struct sk_buff *peek_wr(const struct s3_conn *c3cn) { return c3cn->wr_pending_head; @@ -529,8 +538,8 @@ static inline struct sk_buff *dequeue_wr(struct s3_conn *c3cn) if (likely(skb)) { /* Don't bother clearing the tail */ - c3cn->wr_pending_head = skb_wr_data(skb); - skb_wr_data(skb) = NULL; + c3cn->wr_pending_head = skb_tx_wr_next(skb); + skb_tx_wr_next(skb) = NULL; } return skb; } @@ -543,13 +552,14 @@ static void purge_wr_queue(struct s3_conn *c3cn) } static inline void make_tx_data_wr(struct s3_conn *c3cn, struct sk_buff *skb, - int len) + int len, int req_completion) { struct tx_data_wr *req; skb_reset_transport_header(skb); req = (struct tx_data_wr *)__skb_push(skb, sizeof(*req)); - req->wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_TX_DATA)); + req->wr_hi = htonl(V_WR_OP(FW_WROPCODE_OFLD_TX_DATA) | + (req_completion ? F_WR_COMPL : 0)); req->wr_lo = htonl(V_WR_TID(c3cn->tid)); req->sndseq = htonl(c3cn->snd_nxt); /* len includes the length of any HW ULP additions */ @@ -592,7 +602,7 @@ static int c3cn_push_tx_frames(struct s3_conn *c3cn, int req_completion) if (unlikely(c3cn->state == C3CN_STATE_CONNECTING || c3cn->state == C3CN_STATE_CLOSE_WAIT_1 || - c3cn->state == C3CN_STATE_ABORTING)) { + c3cn->state >= C3CN_STATE_ABORTING)) { c3cn_tx_debug("c3cn 0x%p, in closing state %u.\n", c3cn, c3cn->state); return 0; @@ -615,7 +625,7 @@ static int c3cn_push_tx_frames(struct s3_conn *c3cn, int req_completion) if (c3cn->wr_avail < wrs_needed) { c3cn_tx_debug("c3cn 0x%p, skb len %u/%u, frag %u, " "wr %d < %u.\n", - c3cn, skb->len, skb->datalen, frags, + c3cn, skb->len, skb->data_len, frags, wrs_needed, c3cn->wr_avail); break; } @@ -627,20 +637,24 @@ static int c3cn_push_tx_frames(struct s3_conn *c3cn, int req_completion) c3cn->wr_unacked += wrs_needed; enqueue_wr(c3cn, skb); - if (likely(CXGB3_SKB_CB(skb)->flags & C3CB_FLAG_NEED_HDR)) { - len += ulp_extra_len(skb); - make_tx_data_wr(c3cn, skb, len); - c3cn->snd_nxt += len; - if ((req_completion - && c3cn->wr_unacked == wrs_needed) - || (CXGB3_SKB_CB(skb)->flags & C3CB_FLAG_COMPL) - || c3cn->wr_unacked >= c3cn->wr_max / 2) { - struct work_request_hdr *wr = cplhdr(skb); + c3cn_tx_debug("c3cn 0x%p, enqueue, skb len %u/%u, frag %u, " + "wr %d, left %u, unack %u.\n", + c3cn, skb->len, skb->data_len, frags, + wrs_needed, c3cn->wr_avail, c3cn->wr_unacked); - wr->wr_hi |= htonl(F_WR_COMPL); + + if (likely(skb_flags(skb) & C3CB_FLAG_NEED_HDR)) { + if ((req_completion && + c3cn->wr_unacked == wrs_needed) || + (skb_flags(skb) & C3CB_FLAG_COMPL) || + c3cn->wr_unacked >= c3cn->wr_max / 2) { + req_completion = 1; c3cn->wr_unacked = 0; } - CXGB3_SKB_CB(skb)->flags &= ~C3CB_FLAG_NEED_HDR; + len += ulp_extra_len(skb); + make_tx_data_wr(c3cn, skb, len, req_completion); + c3cn->snd_nxt += len; + skb_flags(skb) &= ~C3CB_FLAG_NEED_HDR; } total_size += skb->truesize; @@ -735,8 +749,11 @@ static void process_act_establish(struct s3_conn *c3cn, struct sk_buff *skb) if (unlikely(c3cn_flag(c3cn, C3CN_ACTIVE_CLOSE_NEEDED))) /* upper layer has requested closing */ send_abort_req(c3cn); - else if (c3cn_push_tx_frames(c3cn, 1)) + else { + if (skb_queue_len(&c3cn->write_queue)) + c3cn_push_tx_frames(c3cn, 1); cxgb3i_conn_tx_open(c3cn); + } } static int do_act_establish(struct t3cdev *cdev, struct sk_buff *skb, @@ -1082,8 +1099,8 @@ static void process_rx_iscsi_hdr(struct s3_conn *c3cn, struct sk_buff *skb) return; } - CXGB3_SKB_CB(skb)->seq = ntohl(hdr_cpl->seq); - CXGB3_SKB_CB(skb)->flags = 0; + skb_tcp_seq(skb) = ntohl(hdr_cpl->seq); + skb_flags(skb) = 0; skb_reset_transport_header(skb); __skb_pull(skb, sizeof(struct cpl_iscsi_hdr)); @@ -1103,12 +1120,12 @@ static void process_rx_iscsi_hdr(struct s3_conn *c3cn, struct sk_buff *skb) goto abort_conn; skb_ulp_mode(skb) = ULP2_FLAG_DATA_READY; - skb_ulp_pdulen(skb) = ntohs(ddp_cpl.len); - skb_ulp_ddigest(skb) = ntohl(ddp_cpl.ulp_crc); + skb_rx_pdulen(skb) = ntohs(ddp_cpl.len); + skb_rx_ddigest(skb) = ntohl(ddp_cpl.ulp_crc); status = ntohl(ddp_cpl.ddp_status); c3cn_rx_debug("rx skb 0x%p, len %u, pdulen %u, ddp status 0x%x.\n", - skb, skb->len, skb_ulp_pdulen(skb), status); + skb, skb->len, skb_rx_pdulen(skb), status); if (status & (1 << RX_DDP_STATUS_HCRC_SHIFT)) skb_ulp_mode(skb) |= ULP2_FLAG_HCRC_ERROR; @@ -1126,7 +1143,7 @@ static void process_rx_iscsi_hdr(struct s3_conn *c3cn, struct sk_buff *skb) } else if (status & (1 << RX_DDP_STATUS_DDP_SHIFT)) skb_ulp_mode(skb) |= ULP2_FLAG_DATA_DDPED; - c3cn->rcv_nxt = ntohl(ddp_cpl.seq) + skb_ulp_pdulen(skb); + c3cn->rcv_nxt = ntohl(ddp_cpl.seq) + skb_rx_pdulen(skb); __pskb_trim(skb, len); __skb_queue_tail(&c3cn->receive_queue, skb); cxgb3i_conn_pdu_ready(c3cn); @@ -1151,12 +1168,27 @@ static int do_iscsi_hdr(struct t3cdev *t3dev, struct sk_buff *skb, void *ctx) * Process an acknowledgment of WR completion. Advance snd_una and send the * next batch of work requests from the write queue. */ +static void check_wr_invariants(struct s3_conn *c3cn) +{ + int pending = count_pending_wrs(c3cn); + + if (unlikely(c3cn->wr_avail + pending != c3cn->wr_max)) + cxgb3i_log_error("TID %u: credit imbalance: avail %u, " + "pending %u, total should be %u\n", + c3cn->tid, c3cn->wr_avail, pending, + c3cn->wr_max); +} + static void process_wr_ack(struct s3_conn *c3cn, struct sk_buff *skb) { struct cpl_wr_ack *hdr = cplhdr(skb); unsigned int credits = ntohs(hdr->credits); u32 snd_una = ntohl(hdr->snd_una); + c3cn_tx_debug("%u WR credits, avail %u, unack %u, TID %u, state %u.\n", + credits, c3cn->wr_avail, c3cn->wr_unacked, + c3cn->tid, c3cn->state); + c3cn->wr_avail += credits; if (c3cn->wr_unacked > c3cn->wr_max - c3cn->wr_avail) c3cn->wr_unacked = c3cn->wr_max - c3cn->wr_avail; @@ -1171,6 +1203,17 @@ static void process_wr_ack(struct s3_conn *c3cn, struct sk_buff *skb) break; } if (unlikely(credits < p->csum)) { + struct tx_data_wr *w = cplhdr(p); + cxgb3i_log_error("TID %u got %u WR credits need %u, " + "len %u, main body %u, frags %u, " + "seq # %u, ACK una %u, ACK nxt %u, " + "WR_AVAIL %u, WRs pending %u\n", + c3cn->tid, credits, p->csum, p->len, + p->len - p->data_len, + skb_shinfo(p)->nr_frags, + ntohl(w->sndseq), snd_una, + ntohl(hdr->snd_nxt), c3cn->wr_avail, + count_pending_wrs(c3cn) - credits); p->csum -= credits; break; } else { @@ -1180,15 +1223,24 @@ static void process_wr_ack(struct s3_conn *c3cn, struct sk_buff *skb) } } - if (unlikely(before(snd_una, c3cn->snd_una))) + check_wr_invariants(c3cn); + + if (unlikely(before(snd_una, c3cn->snd_una))) { + cxgb3i_log_error("TID %u, unexpected sequence # %u in WR_ACK " + "snd_una %u\n", + c3cn->tid, snd_una, c3cn->snd_una); goto out_free; + } if (c3cn->snd_una != snd_una) { c3cn->snd_una = snd_una; dst_confirm(c3cn->dst_cache); } - if (skb_queue_len(&c3cn->write_queue) && c3cn_push_tx_frames(c3cn, 0)) + if (skb_queue_len(&c3cn->write_queue)) { + if (c3cn_push_tx_frames(c3cn, 0)) + cxgb3i_conn_tx_open(c3cn); + } else cxgb3i_conn_tx_open(c3cn); out_free: __kfree_skb(skb); @@ -1452,7 +1504,7 @@ static void init_offload_conn(struct s3_conn *c3cn, struct dst_entry *dst) { BUG_ON(c3cn->cdev != cdev); - c3cn->wr_max = c3cn->wr_avail = T3C_DATA(cdev)->max_wrs; + c3cn->wr_max = c3cn->wr_avail = T3C_DATA(cdev)->max_wrs - 1; c3cn->wr_unacked = 0; c3cn->mss_idx = select_mss(c3cn, dst_mtu(dst)); @@ -1671,9 +1723,17 @@ int cxgb3i_c3cn_send_pdus(struct s3_conn *c3cn, struct sk_buff *skb) goto out_err; } - err = -EPIPE; if (c3cn->err) { c3cn_tx_debug("c3cn 0x%p, err %d.\n", c3cn, c3cn->err); + err = -EPIPE; + goto out_err; + } + + if (c3cn->write_seq - c3cn->snd_una >= cxgb3_snd_win) { + c3cn_tx_debug("c3cn 0x%p, snd %u - %u > %u.\n", + c3cn, c3cn->write_seq, c3cn->snd_una, + cxgb3_snd_win); + err = -EAGAIN; goto out_err; } diff --git a/drivers/scsi/cxgb3i/cxgb3i_offload.h b/drivers/scsi/cxgb3i/cxgb3i_offload.h index d23156907ffd..df1eae0ee4be 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_offload.h +++ b/drivers/scsi/cxgb3i/cxgb3i_offload.h @@ -178,25 +178,33 @@ void cxgb3i_c3cn_release(struct s3_conn *); * @flag: see C3CB_FLAG_* below * @ulp_mode: ULP mode/submode of sk_buff * @seq: tcp sequence number - * @ddigest: pdu data digest - * @pdulen: recovered pdu length - * @wr_data: scratch area for tx wr */ +struct cxgb3_skb_rx_cb { + __u32 ddigest; /* data digest */ + __u32 pdulen; /* recovered pdu length */ +}; + +struct cxgb3_skb_tx_cb { + struct sk_buff *wr_next; /* next wr */ +}; + struct cxgb3_skb_cb { __u8 flags; __u8 ulp_mode; __u32 seq; - __u32 ddigest; - __u32 pdulen; - struct sk_buff *wr_data; + union { + struct cxgb3_skb_rx_cb rx; + struct cxgb3_skb_tx_cb tx; + }; }; #define CXGB3_SKB_CB(skb) ((struct cxgb3_skb_cb *)&((skb)->cb[0])) - +#define skb_flags(skb) (CXGB3_SKB_CB(skb)->flags) #define skb_ulp_mode(skb) (CXGB3_SKB_CB(skb)->ulp_mode) -#define skb_ulp_ddigest(skb) (CXGB3_SKB_CB(skb)->ddigest) -#define skb_ulp_pdulen(skb) (CXGB3_SKB_CB(skb)->pdulen) -#define skb_wr_data(skb) (CXGB3_SKB_CB(skb)->wr_data) +#define skb_tcp_seq(skb) (CXGB3_SKB_CB(skb)->seq) +#define skb_rx_ddigest(skb) (CXGB3_SKB_CB(skb)->rx.ddigest) +#define skb_rx_pdulen(skb) (CXGB3_SKB_CB(skb)->rx.pdulen) +#define skb_tx_wr_next(skb) (CXGB3_SKB_CB(skb)->tx.wr_next) enum c3cb_flags { C3CB_FLAG_NEED_HDR = 1 << 0, /* packet needs a TX_DATA_WR header */ From 949847d195d2bb86f61c289a57edb9207c4a3bbf Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:38:49 -0800 Subject: [PATCH 15/92] [SCSI] cxgb3i: added per-task data to track transmit progress added per-task struct cxgb3i_task_data to track the data transmiting progress and the state of the pdus to be transmitted. Signed-off-by: Karen Xie Reviewed-by: Mike Christie Signed-off-by: James Bottomley --- drivers/scsi/cxgb3i/cxgb3i.h | 21 +++++++++++++++++++++ drivers/scsi/cxgb3i/cxgb3i_iscsi.c | 5 +++-- 2 files changed, 24 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/cxgb3i/cxgb3i.h b/drivers/scsi/cxgb3i/cxgb3i.h index fde6e4c634e7..a7cf550b9cca 100644 --- a/drivers/scsi/cxgb3i/cxgb3i.h +++ b/drivers/scsi/cxgb3i/cxgb3i.h @@ -20,6 +20,7 @@ #include #include #include +#include #include /* from cxgb3 LLD */ @@ -113,6 +114,26 @@ struct cxgb3i_endpoint { struct cxgb3i_conn *cconn; }; +/** + * struct cxgb3i_task_data - private iscsi task data + * + * @nr_frags: # of coalesced page frags (from scsi sgl) + * @frags: coalesced page frags (from scsi sgl) + * @skb: tx pdu skb + * @offset: data offset for the next pdu + * @count: max. possible pdu payload + * @sgoffset: offset to the first sg entry for a given offset + */ +#define MAX_PDU_FRAGS ((ULP2_MAX_PDU_PAYLOAD + 512 - 1) / 512) +struct cxgb3i_task_data { + unsigned short nr_frags; + skb_frag_t frags[MAX_PDU_FRAGS]; + struct sk_buff *skb; + unsigned int offset; + unsigned int count; + unsigned int sgoffset; +}; + int cxgb3i_iscsi_init(void); void cxgb3i_iscsi_cleanup(void); diff --git a/drivers/scsi/cxgb3i/cxgb3i_iscsi.c b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c index d83464b9b3f9..f0434b74554c 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_iscsi.c +++ b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c @@ -364,7 +364,8 @@ cxgb3i_session_create(struct iscsi_endpoint *ep, u16 cmds_max, u16 qdepth, cls_session = iscsi_session_setup(&cxgb3i_iscsi_transport, shost, cmds_max, - sizeof(struct iscsi_tcp_task), + sizeof(struct iscsi_tcp_task) + + sizeof(struct cxgb3i_task_data), initial_cmdsn, ISCSI_MAX_TARGET); if (!cls_session) return NULL; @@ -844,7 +845,7 @@ static struct scsi_host_template cxgb3i_host_template = { .proc_name = "cxgb3i", .queuecommand = iscsi_queuecommand, .change_queue_depth = iscsi_change_queue_depth, - .can_queue = 128 * (ISCSI_DEF_XMIT_CMDS_MAX - 1), + .can_queue = CXGB3I_SCSI_QDEPTH_DFLT - 1, .sg_tablesize = SG_ALL, .max_sectors = 0xFFFF, .cmd_per_lun = ISCSI_DEF_CMD_PER_LUN, From f62d0896e67195d4407ef81c6f77a92f72a63e88 Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:38:54 -0800 Subject: [PATCH 16/92] [SCSI] cxgb3i: Outgoing pdus need to observe skb's MAX_SKB_FRAGS Need to make sure the outgoing pdu can fit into a single skb. When calulating the max. outgoing pdu payload size, take into consideration of - data can be held in the skb's fragment list, assume 512 bytes per fragment, and - data can be held in the headroom. Signed-off-by: Karen Xie Reviewed-by: Mike Christie Signed-off-by: James Bottomley --- drivers/scsi/cxgb3i/cxgb3i_ddp.c | 19 +- drivers/scsi/cxgb3i/cxgb3i_ddp.h | 3 +- drivers/scsi/cxgb3i/cxgb3i_iscsi.c | 17 +- drivers/scsi/cxgb3i/cxgb3i_offload.h | 1 + drivers/scsi/cxgb3i/cxgb3i_pdu.c | 269 ++++++++++++++++++--------- drivers/scsi/cxgb3i/cxgb3i_pdu.h | 2 +- 6 files changed, 202 insertions(+), 109 deletions(-) diff --git a/drivers/scsi/cxgb3i/cxgb3i_ddp.c b/drivers/scsi/cxgb3i/cxgb3i_ddp.c index 08f3a09d9233..a83d36e4926f 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_ddp.c +++ b/drivers/scsi/cxgb3i/cxgb3i_ddp.c @@ -639,10 +639,11 @@ static int ddp_init(struct t3cdev *tdev) write_unlock(&cxgb3i_ddp_rwlock); ddp_log_info("nppods %u (0x%x ~ 0x%x), bits %u, mask 0x%x,0x%x " - "pkt %u,%u.\n", + "pkt %u/%u, %u/%u.\n", ppmax, ddp->llimit, ddp->ulimit, ddp->idx_bits, ddp->idx_mask, ddp->rsvd_tag_mask, - ddp->max_txsz, ddp->max_rxsz); + ddp->max_txsz, uinfo.max_txsz, + ddp->max_rxsz, uinfo.max_rxsz); return 0; free_ddp_map: @@ -654,8 +655,8 @@ free_ddp_map: * cxgb3i_adapter_ddp_init - initialize the adapter's ddp resource * @tdev: t3cdev adapter * @tformat: tag format - * @txsz: max tx pkt size, filled in by this func. - * @rxsz: max rx pkt size, filled in by this func. + * @txsz: max tx pdu payload size, filled in by this func. + * @rxsz: max rx pdu payload size, filled in by this func. * initialize the ddp pagepod manager for a given adapter if needed and * setup the tag format for a given iscsi entity */ @@ -685,10 +686,12 @@ int cxgb3i_adapter_ddp_init(struct t3cdev *tdev, tformat->sw_bits, tformat->rsvd_bits, tformat->rsvd_shift, tformat->rsvd_mask); - *txsz = ddp->max_txsz; - *rxsz = ddp->max_rxsz; - ddp_log_info("ddp max pkt size: %u, %u.\n", - ddp->max_txsz, ddp->max_rxsz); + *txsz = min_t(unsigned int, ULP2_MAX_PDU_PAYLOAD, + ddp->max_txsz - ISCSI_PDU_NONPAYLOAD_LEN); + *rxsz = min_t(unsigned int, ULP2_MAX_PDU_PAYLOAD, + ddp->max_rxsz - ISCSI_PDU_NONPAYLOAD_LEN); + ddp_log_info("max payload size: %u/%u, %u/%u.\n", + *txsz, ddp->max_txsz, *rxsz, ddp->max_rxsz); return 0; } EXPORT_SYMBOL_GPL(cxgb3i_adapter_ddp_init); diff --git a/drivers/scsi/cxgb3i/cxgb3i_ddp.h b/drivers/scsi/cxgb3i/cxgb3i_ddp.h index 5c7c4d95c493..99d7014d614d 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_ddp.h +++ b/drivers/scsi/cxgb3i/cxgb3i_ddp.h @@ -85,8 +85,9 @@ struct cxgb3i_ddp_info { struct sk_buff **gl_skb; }; +#define ISCSI_PDU_NONPAYLOAD_LEN 312 /* bhs(48) + ahs(256) + digest(8) */ #define ULP2_MAX_PKT_SIZE 16224 -#define ULP2_MAX_PDU_PAYLOAD (ULP2_MAX_PKT_SIZE - ISCSI_PDU_NONPAYLOAD_MAX) +#define ULP2_MAX_PDU_PAYLOAD (ULP2_MAX_PKT_SIZE - ISCSI_PDU_NONPAYLOAD_LEN) #define PPOD_PAGES_MAX 4 #define PPOD_PAGES_SHIFT 2 /* 4 pages per pod */ diff --git a/drivers/scsi/cxgb3i/cxgb3i_iscsi.c b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c index f0434b74554c..fa2a44f37b36 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_iscsi.c +++ b/drivers/scsi/cxgb3i/cxgb3i_iscsi.c @@ -403,17 +403,15 @@ static inline int cxgb3i_conn_max_xmit_dlength(struct iscsi_conn *conn) { struct iscsi_tcp_conn *tcp_conn = conn->dd_data; struct cxgb3i_conn *cconn = tcp_conn->dd_data; - unsigned int max = min_t(unsigned int, ULP2_MAX_PDU_PAYLOAD, - cconn->hba->snic->tx_max_size - - ISCSI_PDU_NONPAYLOAD_MAX); + unsigned int max = max(512 * MAX_SKB_FRAGS, SKB_TX_HEADROOM); + max = min(cconn->hba->snic->tx_max_size, max); if (conn->max_xmit_dlength) - conn->max_xmit_dlength = min_t(unsigned int, - conn->max_xmit_dlength, max); + conn->max_xmit_dlength = min(conn->max_xmit_dlength, max); else conn->max_xmit_dlength = max; align_pdu_size(conn->max_xmit_dlength); - cxgb3i_log_info("conn 0x%p, max xmit %u.\n", + cxgb3i_api_debug("conn 0x%p, max xmit %u.\n", conn, conn->max_xmit_dlength); return 0; } @@ -428,9 +426,7 @@ static inline int cxgb3i_conn_max_recv_dlength(struct iscsi_conn *conn) { struct iscsi_tcp_conn *tcp_conn = conn->dd_data; struct cxgb3i_conn *cconn = tcp_conn->dd_data; - unsigned int max = min_t(unsigned int, ULP2_MAX_PDU_PAYLOAD, - cconn->hba->snic->rx_max_size - - ISCSI_PDU_NONPAYLOAD_MAX); + unsigned int max = cconn->hba->snic->rx_max_size; align_pdu_size(max); if (conn->max_recv_dlength) { @@ -440,8 +436,7 @@ static inline int cxgb3i_conn_max_recv_dlength(struct iscsi_conn *conn) conn->max_recv_dlength, max); return -EINVAL; } - conn->max_recv_dlength = min_t(unsigned int, - conn->max_recv_dlength, max); + conn->max_recv_dlength = min(conn->max_recv_dlength, max); align_pdu_size(conn->max_recv_dlength); } else conn->max_recv_dlength = max; diff --git a/drivers/scsi/cxgb3i/cxgb3i_offload.h b/drivers/scsi/cxgb3i/cxgb3i_offload.h index df1eae0ee4be..6344b9eb2589 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_offload.h +++ b/drivers/scsi/cxgb3i/cxgb3i_offload.h @@ -225,6 +225,7 @@ struct sge_opaque_hdr { /* for TX: a skb must have a headroom of at least TX_HEADER_LEN bytes */ #define TX_HEADER_LEN \ (sizeof(struct tx_data_wr) + sizeof(struct sge_opaque_hdr)) +#define SKB_TX_HEADROOM SKB_MAX_HEAD(TX_HEADER_LEN) /* * get and set private ip for iscsi traffic diff --git a/drivers/scsi/cxgb3i/cxgb3i_pdu.c b/drivers/scsi/cxgb3i/cxgb3i_pdu.c index ce7ce8c6094c..17115c230d65 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_pdu.c +++ b/drivers/scsi/cxgb3i/cxgb3i_pdu.c @@ -32,6 +32,10 @@ #define cxgb3i_tx_debug(fmt...) #endif +/* always allocate rooms for AHS */ +#define SKB_TX_PDU_HEADER_LEN \ + (sizeof(struct iscsi_hdr) + ISCSI_MAX_AHS_SIZE) +static unsigned int skb_extra_headroom; static struct page *pad_page; /* @@ -146,12 +150,13 @@ static inline void tx_skb_setmode(struct sk_buff *skb, int hcrc, int dcrc) void cxgb3i_conn_cleanup_task(struct iscsi_task *task) { - struct iscsi_tcp_task *tcp_task = task->dd_data; + struct cxgb3i_task_data *tdata = task->dd_data + + sizeof(struct iscsi_tcp_task); /* never reached the xmit task callout */ - if (tcp_task->dd_data) - kfree_skb(tcp_task->dd_data); - tcp_task->dd_data = NULL; + if (tdata->skb) + __kfree_skb(tdata->skb); + memset(tdata, 0, sizeof(struct cxgb3i_task_data)); /* MNC - Do we need a check in case this is called but * cxgb3i_conn_alloc_pdu has never been called on the task */ @@ -159,28 +164,102 @@ void cxgb3i_conn_cleanup_task(struct iscsi_task *task) iscsi_tcp_cleanup_task(task); } -/* - * We do not support ahs yet - */ +static int sgl_seek_offset(struct scatterlist *sgl, unsigned int sgcnt, + unsigned int offset, unsigned int *off, + struct scatterlist **sgp) +{ + int i; + struct scatterlist *sg; + + for_each_sg(sgl, sg, sgcnt, i) { + if (offset < sg->length) { + *off = offset; + *sgp = sg; + return 0; + } + offset -= sg->length; + } + return -EFAULT; +} + +static int sgl_read_to_frags(struct scatterlist *sg, unsigned int sgoffset, + unsigned int dlen, skb_frag_t *frags, + int frag_max) +{ + unsigned int datalen = dlen; + unsigned int sglen = sg->length - sgoffset; + struct page *page = sg_page(sg); + int i; + + i = 0; + do { + unsigned int copy; + + if (!sglen) { + sg = sg_next(sg); + if (!sg) { + cxgb3i_log_error("%s, sg NULL, len %u/%u.\n", + __func__, datalen, dlen); + return -EINVAL; + } + sgoffset = 0; + sglen = sg->length; + page = sg_page(sg); + + } + copy = min(datalen, sglen); + if (i && page == frags[i - 1].page && + sgoffset + sg->offset == + frags[i - 1].page_offset + frags[i - 1].size) { + frags[i - 1].size += copy; + } else { + if (i >= frag_max) { + cxgb3i_log_error("%s, too many pages %u, " + "dlen %u.\n", __func__, + frag_max, dlen); + return -EINVAL; + } + + frags[i].page = page; + frags[i].page_offset = sg->offset + sgoffset; + frags[i].size = copy; + i++; + } + datalen -= copy; + sgoffset += copy; + sglen -= copy; + } while (datalen); + + return i; +} + int cxgb3i_conn_alloc_pdu(struct iscsi_task *task, u8 opcode) { + struct iscsi_conn *conn = task->conn; struct iscsi_tcp_task *tcp_task = task->dd_data; - struct sk_buff *skb; + struct cxgb3i_task_data *tdata = task->dd_data + sizeof(*tcp_task); + struct scsi_cmnd *sc = task->sc; + int headroom = SKB_TX_PDU_HEADER_LEN; + tcp_task->dd_data = tdata; task->hdr = NULL; - /* always allocate rooms for AHS */ - skb = alloc_skb(sizeof(struct iscsi_hdr) + ISCSI_MAX_AHS_SIZE + - TX_HEADER_LEN, GFP_ATOMIC); - if (!skb) + + /* write command, need to send data pdus */ + if (skb_extra_headroom && (opcode == ISCSI_OP_SCSI_DATA_OUT || + (opcode == ISCSI_OP_SCSI_CMD && + (scsi_bidi_cmnd(sc) || sc->sc_data_direction == DMA_TO_DEVICE)))) + headroom += min(skb_extra_headroom, conn->max_xmit_dlength); + + tdata->skb = alloc_skb(TX_HEADER_LEN + headroom, GFP_ATOMIC); + if (!tdata->skb) return -ENOMEM; + skb_reserve(tdata->skb, TX_HEADER_LEN); cxgb3i_tx_debug("task 0x%p, opcode 0x%x, skb 0x%p.\n", - task, opcode, skb); + task, opcode, tdata->skb); - tcp_task->dd_data = skb; - skb_reserve(skb, TX_HEADER_LEN); - task->hdr = (struct iscsi_hdr *)skb->data; - task->hdr_max = sizeof(struct iscsi_hdr); + task->hdr = (struct iscsi_hdr *)tdata->skb->data; + task->hdr_max = SKB_TX_PDU_HEADER_LEN; /* data_out uses scsi_cmd's itt */ if (opcode != ISCSI_OP_SCSI_DATA_OUT) @@ -192,13 +271,13 @@ int cxgb3i_conn_alloc_pdu(struct iscsi_task *task, u8 opcode) int cxgb3i_conn_init_pdu(struct iscsi_task *task, unsigned int offset, unsigned int count) { - struct iscsi_tcp_task *tcp_task = task->dd_data; - struct sk_buff *skb = tcp_task->dd_data; struct iscsi_conn *conn = task->conn; - struct page *pg; + struct iscsi_tcp_task *tcp_task = task->dd_data; + struct cxgb3i_task_data *tdata = tcp_task->dd_data; + struct sk_buff *skb = tdata->skb; unsigned int datalen = count; int i, padlen = iscsi_padding(count); - skb_frag_t *frag; + struct page *pg; cxgb3i_tx_debug("task 0x%p,0x%p, offset %u, count %u, skb 0x%p.\n", task, task->sc, offset, count, skb); @@ -209,90 +288,94 @@ int cxgb3i_conn_init_pdu(struct iscsi_task *task, unsigned int offset, return 0; if (task->sc) { - struct scatterlist *sg; - struct scsi_data_buffer *sdb; - unsigned int sgoffset = offset; - struct page *sgpg; - unsigned int sglen; + struct scsi_data_buffer *sdb = scsi_out(task->sc); + struct scatterlist *sg = NULL; + int err; - sdb = scsi_out(task->sc); - sg = sdb->table.sgl; - - for_each_sg(sdb->table.sgl, sg, sdb->table.nents, i) { - cxgb3i_tx_debug("sg %d, page 0x%p, len %u offset %u\n", - i, sg_page(sg), sg->length, sg->offset); - - if (sgoffset < sg->length) - break; - sgoffset -= sg->length; + tdata->offset = offset; + tdata->count = count; + err = sgl_seek_offset(sdb->table.sgl, sdb->table.nents, + tdata->offset, &tdata->sgoffset, &sg); + if (err < 0) { + cxgb3i_log_warn("tpdu, sgl %u, bad offset %u/%u.\n", + sdb->table.nents, tdata->offset, + sdb->length); + return err; } - sgpg = sg_page(sg); - sglen = sg->length - sgoffset; + err = sgl_read_to_frags(sg, tdata->sgoffset, tdata->count, + tdata->frags, MAX_PDU_FRAGS); + if (err < 0) { + cxgb3i_log_warn("tpdu, sgl %u, bad offset %u + %u.\n", + sdb->table.nents, tdata->offset, + tdata->count); + return err; + } + tdata->nr_frags = err; - do { - int j = skb_shinfo(skb)->nr_frags; - unsigned int copy; + if (tdata->nr_frags > MAX_SKB_FRAGS || + (padlen && tdata->nr_frags == MAX_SKB_FRAGS)) { + char *dst = skb->data + task->hdr_len; + skb_frag_t *frag = tdata->frags; - if (!sglen) { - sg = sg_next(sg); - sgpg = sg_page(sg); - sgoffset = 0; - sglen = sg->length; - ++i; + /* data fits in the skb's headroom */ + for (i = 0; i < tdata->nr_frags; i++, frag++) { + char *src = kmap_atomic(frag->page, + KM_SOFTIRQ0); + + memcpy(dst, src+frag->page_offset, frag->size); + dst += frag->size; + kunmap_atomic(src, KM_SOFTIRQ0); } - copy = min(sglen, datalen); - if (j && skb_can_coalesce(skb, j, sgpg, - sg->offset + sgoffset)) { - skb_shinfo(skb)->frags[j - 1].size += copy; - } else { - get_page(sgpg); - skb_fill_page_desc(skb, j, sgpg, - sg->offset + sgoffset, copy); + if (padlen) { + memset(dst, 0, padlen); + padlen = 0; } - sgoffset += copy; - sglen -= copy; - datalen -= copy; - } while (datalen); + skb_put(skb, count + padlen); + } else { + /* data fit into frag_list */ + for (i = 0; i < tdata->nr_frags; i++) + get_page(tdata->frags[i].page); + + memcpy(skb_shinfo(skb)->frags, tdata->frags, + sizeof(skb_frag_t) * tdata->nr_frags); + skb_shinfo(skb)->nr_frags = tdata->nr_frags; + skb->len += count; + skb->data_len += count; + skb->truesize += count; + } + } else { pg = virt_to_page(task->data); - while (datalen) { - i = skb_shinfo(skb)->nr_frags; - frag = &skb_shinfo(skb)->frags[i]; - - get_page(pg); - frag->page = pg; - frag->page_offset = 0; - frag->size = min((unsigned int)PAGE_SIZE, datalen); - - skb_shinfo(skb)->nr_frags++; - datalen -= frag->size; - pg++; - } + get_page(pg); + skb_fill_page_desc(skb, 0, pg, offset_in_page(task->data), + count); + skb->len += count; + skb->data_len += count; + skb->truesize += count; } if (padlen) { i = skb_shinfo(skb)->nr_frags; - frag = &skb_shinfo(skb)->frags[i]; - frag->page = pad_page; - frag->page_offset = 0; - frag->size = padlen; - skb_shinfo(skb)->nr_frags++; + get_page(pad_page); + skb_fill_page_desc(skb, skb_shinfo(skb)->nr_frags, pad_page, 0, + padlen); + + skb->data_len += padlen; + skb->truesize += padlen; + skb->len += padlen; } - datalen = count + padlen; - skb->data_len += datalen; - skb->truesize += datalen; - skb->len += datalen; return 0; } int cxgb3i_conn_xmit_pdu(struct iscsi_task *task) { - struct iscsi_tcp_task *tcp_task = task->dd_data; - struct sk_buff *skb = tcp_task->dd_data; struct iscsi_tcp_conn *tcp_conn = task->conn->dd_data; struct cxgb3i_conn *cconn = tcp_conn->dd_data; + struct iscsi_tcp_task *tcp_task = task->dd_data; + struct cxgb3i_task_data *tdata = tcp_task->dd_data; + struct sk_buff *skb = tdata->skb; unsigned int datalen; int err; @@ -300,13 +383,14 @@ int cxgb3i_conn_xmit_pdu(struct iscsi_task *task) return 0; datalen = skb->data_len; - tcp_task->dd_data = NULL; + tdata->skb = NULL; err = cxgb3i_c3cn_send_pdus(cconn->cep->c3cn, skb); - cxgb3i_tx_debug("task 0x%p, skb 0x%p, len %u/%u, rv %d.\n", - task, skb, skb->len, skb->data_len, err); if (err > 0) { int pdulen = err; + cxgb3i_tx_debug("task 0x%p, skb 0x%p, len %u/%u, rv %d.\n", + task, skb, skb->len, skb->data_len, err); + if (task->conn->hdrdgst_en) pdulen += ISCSI_DIGEST_SIZE; if (datalen && task->conn->datadgst_en) @@ -325,12 +409,14 @@ int cxgb3i_conn_xmit_pdu(struct iscsi_task *task) return err; } /* reset skb to send when we are called again */ - tcp_task->dd_data = skb; + tdata->skb = skb; return -EAGAIN; } int cxgb3i_pdu_init(void) { + if (SKB_TX_HEADROOM > (512 * MAX_SKB_FRAGS)) + skb_extra_headroom = SKB_TX_HEADROOM; pad_page = alloc_page(GFP_KERNEL); if (!pad_page) return -ENOMEM; @@ -366,7 +452,9 @@ void cxgb3i_conn_pdu_ready(struct s3_conn *c3cn) skb = skb_peek(&c3cn->receive_queue); while (!err && skb) { __skb_unlink(skb, &c3cn->receive_queue); - read += skb_ulp_pdulen(skb); + read += skb_rx_pdulen(skb); + cxgb3i_rx_debug("conn 0x%p, cn 0x%p, rx skb 0x%p, pdulen %u.\n", + conn, c3cn, skb, skb_rx_pdulen(skb)); err = cxgb3i_conn_read_pdu_skb(conn, skb); __kfree_skb(skb); skb = skb_peek(&c3cn->receive_queue); @@ -377,6 +465,11 @@ void cxgb3i_conn_pdu_ready(struct s3_conn *c3cn) cxgb3i_c3cn_rx_credits(c3cn, read); } conn->rxdata_octets += read; + + if (err) { + cxgb3i_log_info("conn 0x%p rx failed err %d.\n", conn, err); + iscsi_conn_failure(conn, ISCSI_ERR_CONN_FAILED); + } } void cxgb3i_conn_tx_open(struct s3_conn *c3cn) diff --git a/drivers/scsi/cxgb3i/cxgb3i_pdu.h b/drivers/scsi/cxgb3i/cxgb3i_pdu.h index a3f685cc2362..0770b23d90da 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_pdu.h +++ b/drivers/scsi/cxgb3i/cxgb3i_pdu.h @@ -53,7 +53,7 @@ struct cpl_rx_data_ddp_norss { #define ULP2_FLAG_DCRC_ERROR 0x20 #define ULP2_FLAG_PAD_ERROR 0x40 -void cxgb3i_conn_closing(struct s3_conn *); +void cxgb3i_conn_closing(struct s3_conn *c3cn); void cxgb3i_conn_pdu_ready(struct s3_conn *c3cn); void cxgb3i_conn_tx_open(struct s3_conn *c3cn); #endif From 992040f54069c96a59343976950f174448f4a351 Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:38:59 -0800 Subject: [PATCH 17/92] [SCSI] cxgb3i: added missing include in cxgb3i_ddp.h Signed-off-by: Karen Xie Reviewed-by: Mike Christie Signed-off-by: James Bottomley --- drivers/scsi/cxgb3i/cxgb3i_ddp.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/scsi/cxgb3i/cxgb3i_ddp.h b/drivers/scsi/cxgb3i/cxgb3i_ddp.h index 99d7014d614d..3faae7831c83 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_ddp.h +++ b/drivers/scsi/cxgb3i/cxgb3i_ddp.h @@ -13,6 +13,8 @@ #ifndef __CXGB3I_ULP2_DDP_H__ #define __CXGB3I_ULP2_DDP_H__ +#include + /** * struct cxgb3i_tag_format - cxgb3i ulp tag format for an iscsi entity * From 802d52734adf0f288c49c05ed433872d7559c932 Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:39:04 -0800 Subject: [PATCH 18/92] [SCSI] cxgb3i: Fix spelling errors in documentation Signed-off-by: Karen Xie Signed-off-by: James Bottomley --- Documentation/scsi/cxgb3i.txt | 11 +++++------ 1 file changed, 5 insertions(+), 6 deletions(-) diff --git a/Documentation/scsi/cxgb3i.txt b/Documentation/scsi/cxgb3i.txt index 8141fa01978e..7ac8032ee9b2 100644 --- a/Documentation/scsi/cxgb3i.txt +++ b/Documentation/scsi/cxgb3i.txt @@ -4,7 +4,7 @@ Introduction ============ The Chelsio T3 ASIC based Adapters (S310, S320, S302, S304, Mezz cards, etc. -series of products) supports iSCSI acceleration and iSCSI Direct Data Placement +series of products) support iSCSI acceleration and iSCSI Direct Data Placement (DDP) where the hardware handles the expensive byte touching operations, such as CRC computation and verification, and direct DMA to the final host memory destination: @@ -31,9 +31,9 @@ destination: the TCP segments onto the wire. It handles TCP retransmission if needed. - On receving, S3 h/w recovers the iSCSI PDU by reassembling TCP + On receiving, S3 h/w recovers the iSCSI PDU by reassembling TCP segments, separating the header and data, calculating and verifying - the digests, then forwards the header to the host. The payload data, + the digests, then forwarding the header to the host. The payload data, if possible, will be directly placed into the pre-posted host DDP buffer. Otherwise, the payload data will be sent to the host too. @@ -68,9 +68,8 @@ The following steps need to be taken to accelerates the open-iscsi initiator: sure the ip address is unique in the network. 3. edit /etc/iscsi/iscsid.conf - The default setting for MaxRecvDataSegmentLength (131072) is too big, - replace "node.conn[0].iscsi.MaxRecvDataSegmentLength" to be a value no - bigger than 15360 (for example 8192): + The default setting for MaxRecvDataSegmentLength (131072) is too big; + replace with a value no bigger than 15360 (for example 8192): node.conn[0].iscsi.MaxRecvDataSegmentLength = 8192 From b7e7bd34465518f3527bf47a8055f35077d40c6c Mon Sep 17 00:00:00 2001 From: Karen Xie Date: Fri, 13 Feb 2009 21:39:09 -0800 Subject: [PATCH 19/92] [SCSI] cxgb3i: update the driver version to 1.0.1 Signed-off-by: Karen Xie Signed-off-by: James Bottomley --- drivers/scsi/cxgb3i/cxgb3i_init.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/scsi/cxgb3i/cxgb3i_init.c b/drivers/scsi/cxgb3i/cxgb3i_init.c index 091ecb4d9f3d..1ce9f244e46c 100644 --- a/drivers/scsi/cxgb3i/cxgb3i_init.c +++ b/drivers/scsi/cxgb3i/cxgb3i_init.c @@ -12,8 +12,8 @@ #include "cxgb3i.h" #define DRV_MODULE_NAME "cxgb3i" -#define DRV_MODULE_VERSION "1.0.0" -#define DRV_MODULE_RELDATE "Jun. 1, 2008" +#define DRV_MODULE_VERSION "1.0.1" +#define DRV_MODULE_RELDATE "Jan. 2009" static char version[] = "Chelsio S3xx iSCSI Driver " DRV_MODULE_NAME From 4034cc68157bfa0b6622efe368488d3d3e20f4e6 Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Sat, 21 Feb 2009 11:04:45 +0900 Subject: [PATCH 20/92] [SCSI] sd: revive sd_index_lock Commit f27bac2761cab5a2e212dea602d22457a9aa6943 which converted sd to use ida instead of idr incorrectly removed sd_index_lock around id allocation and free. idr/ida do have internal locks but they protect their free object lists not the allocation itself. The caller is responsible for that. This missing synchronization led to the same id being assigned to multiple devices leading to oops. Reported and tracked down by Stuart Hayes of Dell. Signed-off-by: Tejun Heo Cc: Stable Tree Signed-off-by: James Bottomley --- drivers/scsi/sd.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/drivers/scsi/sd.c b/drivers/scsi/sd.c index d57566b8be0a..55310dbc10a6 100644 --- a/drivers/scsi/sd.c +++ b/drivers/scsi/sd.c @@ -107,6 +107,7 @@ static void scsi_disk_release(struct device *cdev); static void sd_print_sense_hdr(struct scsi_disk *, struct scsi_sense_hdr *); static void sd_print_result(struct scsi_disk *, int); +static DEFINE_SPINLOCK(sd_index_lock); static DEFINE_IDA(sd_index_ida); /* This semaphore is used to mediate the 0->1 reference get in the @@ -1914,7 +1915,9 @@ static int sd_probe(struct device *dev) if (!ida_pre_get(&sd_index_ida, GFP_KERNEL)) goto out_put; + spin_lock(&sd_index_lock); error = ida_get_new(&sd_index_ida, &index); + spin_unlock(&sd_index_lock); } while (error == -EAGAIN); if (error) @@ -1936,7 +1939,9 @@ static int sd_probe(struct device *dev) return 0; out_free_index: + spin_lock(&sd_index_lock); ida_remove(&sd_index_ida, index); + spin_unlock(&sd_index_lock); out_put: put_disk(gd); out_free: @@ -1986,7 +1991,9 @@ static void scsi_disk_release(struct device *dev) struct scsi_disk *sdkp = to_scsi_disk(dev); struct gendisk *disk = sdkp->disk; + spin_lock(&sd_index_lock); ida_remove(&sd_index_ida, sdkp->index); + spin_unlock(&sd_index_lock); disk->private_data = NULL; put_disk(disk); From 126c098296c8f96cf7f6ca0fdb47265ac7994f00 Mon Sep 17 00:00:00 2001 From: James Bottomley Date: Thu, 19 Feb 2009 21:48:54 +0000 Subject: [PATCH 21/92] [SCSI] fix ABORTED_COMMAND looping forever problem Instead of terminating after five retries, commands terminated by ABORTED_COMMAND sense are retrying forever. The problem was introduced by: commit b60af5b0adf0da24c673598c8d3fb4d4189a15ce Author: Alan Stern Date: Mon Nov 3 15:56:47 2008 -0500 [SCSI] simplify scsi_io_completion() Which introduced an error whereby ABORTED_COMMAND now gets erroneously retried in scsi_io_completion. Fix this by returning the behaviour back to the default no retry. Reported-by: Sitsofe Wheeler Tested-by: Sitsofe Wheeler Signed-off-by: James Bottomley --- drivers/scsi/scsi_lib.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c index 940dc32ff0dc..b82ffd90632e 100644 --- a/drivers/scsi/scsi_lib.c +++ b/drivers/scsi/scsi_lib.c @@ -1040,12 +1040,11 @@ void scsi_io_completion(struct scsi_cmnd *cmd, unsigned int good_bytes) action = ACTION_FAIL; break; case ABORTED_COMMAND: + action = ACTION_FAIL; if (sshdr.asc == 0x10) { /* DIF */ description = "Target Data Integrity Failure"; - action = ACTION_FAIL; error = -EILSEQ; - } else - action = ACTION_RETRY; + } break; case NOT_READY: /* If the device is in the process of becoming From 8cfd9e923be54ef66ce174a93f4592b444b96407 Mon Sep 17 00:00:00 2001 From: Russell King Date: Sun, 22 Feb 2009 12:36:55 +0000 Subject: [PATCH 22/92] [ARM] RiscPC: Fix etherh oops The 8390 driver was structured by Al Viro to allow the flexibility required by platforms. lib8390.c contains the core code which drivers explicitly include: - 8390.c includes lib8390.c to provide the standard ISA based driver. - etherh.c includes it with the accessors defined for RiscPC platforms, where it is addressed via the MMIO accessors with a device dependent register spacing. Other platform drivers do something similar. However, b9a9b4b caused the kernel to contain not only the etherh private build of lib8390 (included in etherh.c) but also lib8390.c itself, and referred the new net_device_ops methods to the ISA version. The result of this is is not pretty: Unable to handle kernel paging request at virtual address 12032030 pgd = c8330000 [12032030] *pgd=00000000 Internal error: Oops: 18331805 [#1] Modules linked in: ipv6 CPU: 0 Not tainted (2.6.29-rc3 #167) PC is at do_set_multicast_list+0xd0/0x190 LR is at bitrev32+0x28/0x34 pc : [] lr : [] psr: a0000093 sp : c8321d9c ip : c8321d84 fp : c8321dbc r10: c80c6800 r9 : 00000000 r8 : c80c6b60 r7 : c80c6b80 r6 : cc80c800 r5 : c80c6800 r4 : 00000000 r3 : cc80c80c r2 : 00000004 r1 : 00000007 r0 : e0000000 Flags: NzCv IRQs off FIQs on Mode SVC_32 ISA ARM Segment user ... Fix up b9a9b4b by making etherh's net_device_ops refer to the internal lib8390 functions, and remove the build of the ISA 8390.c driver. Acked-by: David S. Miller Signed-off-by: Russell King --- drivers/net/arm/Makefile | 2 +- drivers/net/arm/etherh.c | 10 +++++----- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/net/arm/Makefile b/drivers/net/arm/Makefile index c69c0cdba4a2..811a3ccd14c1 100644 --- a/drivers/net/arm/Makefile +++ b/drivers/net/arm/Makefile @@ -4,7 +4,7 @@ # obj-$(CONFIG_ARM_AM79C961A) += am79c961a.o -obj-$(CONFIG_ARM_ETHERH) += etherh.o ../8390.o +obj-$(CONFIG_ARM_ETHERH) += etherh.o obj-$(CONFIG_ARM_ETHER3) += ether3.o obj-$(CONFIG_ARM_ETHER1) += ether1.o obj-$(CONFIG_ARM_AT91_ETHER) += at91_ether.o diff --git a/drivers/net/arm/etherh.c b/drivers/net/arm/etherh.c index 54b52e5b1821..f52f668c49bf 100644 --- a/drivers/net/arm/etherh.c +++ b/drivers/net/arm/etherh.c @@ -641,15 +641,15 @@ static const struct net_device_ops etherh_netdev_ops = { .ndo_open = etherh_open, .ndo_stop = etherh_close, .ndo_set_config = etherh_set_config, - .ndo_start_xmit = ei_start_xmit, - .ndo_tx_timeout = ei_tx_timeout, - .ndo_get_stats = ei_get_stats, - .ndo_set_multicast_list = ei_set_multicast_list, + .ndo_start_xmit = __ei_start_xmit, + .ndo_tx_timeout = __ei_tx_timeout, + .ndo_get_stats = __ei_get_stats, + .ndo_set_multicast_list = __ei_set_multicast_list, .ndo_validate_addr = eth_validate_addr, .ndo_set_mac_address = eth_mac_addr, .ndo_change_mtu = eth_change_mtu, #ifdef CONFIG_NET_POLL_CONTROLLER - .ndo_poll_controller = ei_poll, + .ndo_poll_controller = __ei_poll, #endif }; From 5ce7868e159a3ee4ddf95f8522643991fea97cf2 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Wed, 18 Feb 2009 11:25:32 -0800 Subject: [PATCH 23/92] [SCSI] mpt: fix disable lsi sas to use msi as default Impact: fix bug the third param in module_param(,,) is perm instead of default value. we still need to assign default at first. Also, the default is now zero not one, so fix the parameter text to reflect that. Signed-off-by: Yinghai Lu Acked-by: Kashyap Desai Signed-off-by: James Bottomley --- drivers/message/fusion/mptbase.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/message/fusion/mptbase.c b/drivers/message/fusion/mptbase.c index 96ac88317b8e..ea3aafbbda44 100644 --- a/drivers/message/fusion/mptbase.c +++ b/drivers/message/fusion/mptbase.c @@ -91,9 +91,9 @@ MODULE_PARM_DESC(mpt_msi_enable_fc, " Enable MSI Support for FC \ controllers (default=0)"); static int mpt_msi_enable_sas; -module_param(mpt_msi_enable_sas, int, 1); +module_param(mpt_msi_enable_sas, int, 0); MODULE_PARM_DESC(mpt_msi_enable_sas, " Enable MSI Support for SAS \ - controllers (default=1)"); + controllers (default=0)"); static int mpt_channel_mapping; From d82ad6d6833ee8248e6c34e1b9fc3c4e61e3f035 Mon Sep 17 00:00:00 2001 From: Andrei Birjukov Date: Sun, 22 Feb 2009 22:37:21 +0000 Subject: [PATCH 24/92] [ARM] at91: fix for Atmel AT91 powersaving We've discovered that our AT91SAM9260 board consumed too much power when returning from a slowclock low-power mode. RAM self-refresh is enabled in a bootloader in our case, this is how we saw a difference. Estimated ca. 30mA more on 4V battery than the same state before powersaving. After a small research we found that there seems to be a bogus sdram_selfrefresh_disable() call at the end of at91_pm_enter() call, which overwrites the LPR register with uninitialized value. Please find the suggested patch attached. This patch fixes correct restoring of LPR register of the Atmel AT91 SDRAM controller when returning from a power saving mode. Signed-off-by: Andrei Birjukov Acked-by: Andrew Victor Signed-off-by: Andrew Morton Signed-off-by: Russell King --- arch/arm/mach-at91/pm.c | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/arm/mach-at91/pm.c b/arch/arm/mach-at91/pm.c index 9bb4f043aa22..7ac812dc055a 100644 --- a/arch/arm/mach-at91/pm.c +++ b/arch/arm/mach-at91/pm.c @@ -332,7 +332,6 @@ static int at91_pm_enter(suspend_state_t state) at91_sys_read(AT91_AIC_IPR) & at91_sys_read(AT91_AIC_IMR)); error: - sdram_selfrefresh_disable(); target_state = PM_SUSPEND_ON; at91_irq_resume(); at91_gpio_resume(); From c8532db7f2661b63f658b9a08cf4053a3e6abb78 Mon Sep 17 00:00:00 2001 From: Mark Brown Date: Tue, 24 Feb 2009 15:55:48 +0100 Subject: [PATCH 25/92] [ARM] 5411/1: S3C64XX: Fix EINT unmask Currently the unmask function for EINT interrupts was setting the mask bit rather than clearing it. This was also previously reported and fixed by Kyungmin Park and others. Acked-By: Ben Dooks Signed-off-by: Mark Brown Signed-off-by: Russell King --- arch/arm/plat-s3c64xx/irq-eint.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/arm/plat-s3c64xx/irq-eint.c b/arch/arm/plat-s3c64xx/irq-eint.c index 1f7cc0067f5c..ebb305ce7689 100644 --- a/arch/arm/plat-s3c64xx/irq-eint.c +++ b/arch/arm/plat-s3c64xx/irq-eint.c @@ -55,7 +55,7 @@ static void s3c_irq_eint_unmask(unsigned int irq) u32 mask; mask = __raw_readl(S3C64XX_EINT0MASK); - mask |= eint_irq_to_bit(irq); + mask &= ~eint_irq_to_bit(irq); __raw_writel(mask, S3C64XX_EINT0MASK); } From a682604838763981613e42015cd0e39f2989d6bb Mon Sep 17 00:00:00 2001 From: "Paul E. McKenney" Date: Wed, 25 Feb 2009 18:03:42 -0800 Subject: [PATCH 26/92] rcu: Teach RCU that idle task is not quiscent state at boot This patch fixes a bug located by Vegard Nossum with the aid of kmemcheck, updated based on review comments from Nick Piggin, Ingo Molnar, and Andrew Morton. And cleans up the variable-name and function-name language. ;-) The boot CPU runs in the context of its idle thread during boot-up. During this time, idle_cpu(0) will always return nonzero, which will fool Classic and Hierarchical RCU into deciding that a large chunk of the boot-up sequence is a big long quiescent state. This in turn causes RCU to prematurely end grace periods during this time. This patch changes the rcutree.c and rcuclassic.c rcu_check_callbacks() function to ignore the idle task as a quiescent state until the system has started up the scheduler in rest_init(), introducing a new non-API function rcu_idle_now_means_idle() to inform RCU of this transition. RCU maintains an internal rcu_idle_cpu_truthful variable to track this state, which is then used by rcu_check_callback() to determine if it should believe idle_cpu(). Because this patch has the effect of disallowing RCU grace periods during long stretches of the boot-up sequence, this patch also introduces Josh Triplett's UP-only optimization that makes synchronize_rcu() be a no-op if num_online_cpus() returns 1. This allows boot-time code that calls synchronize_rcu() to proceed normally. Note, however, that RCU callbacks registered by call_rcu() will likely queue up until later in the boot sequence. Although rcuclassic and rcutree can also use this same optimization after boot completes, rcupreempt must restrict its use of this optimization to the portion of the boot sequence before the scheduler starts up, given that an rcupreempt RCU read-side critical section may be preeempted. In addition, this patch takes Nick Piggin's suggestion to make the system_state global variable be __read_mostly. Changes since v4: o Changes the name of the introduced function and variable to be less emotional. ;-) Changes since v3: o WARN_ON(nr_context_switches() > 0) to verify that RCU switches out of boot-time mode before the first context switch, as suggested by Nick Piggin. Changes since v2: o Created rcu_blocking_is_gp() internal-to-RCU API that determines whether a call to synchronize_rcu() is itself a grace period. o The definition of rcu_blocking_is_gp() for rcuclassic and rcutree checks to see if but a single CPU is online. o The definition of rcu_blocking_is_gp() for rcupreempt checks to see both if but a single CPU is online and if the system is still in early boot. This allows rcupreempt to again work correctly if running on a single CPU after booting is complete. o Added check to rcupreempt's synchronize_sched() for there being but one online CPU. Tested all three variants both SMP and !SMP, booted fine, passed a short rcutorture test on both x86 and Power. Located-by: Vegard Nossum Tested-by: Vegard Nossum Tested-by: Paul E. McKenney Signed-off-by: Paul E. McKenney Signed-off-by: Ingo Molnar --- include/linux/rcuclassic.h | 6 ++++++ include/linux/rcupdate.h | 4 ++++ include/linux/rcupreempt.h | 15 +++++++++++++++ include/linux/rcutree.h | 6 ++++++ init/main.c | 3 ++- kernel/rcuclassic.c | 4 ++-- kernel/rcupdate.c | 12 ++++++++++++ kernel/rcupreempt.c | 3 +++ kernel/rcutree.c | 4 ++-- 9 files changed, 52 insertions(+), 5 deletions(-) diff --git a/include/linux/rcuclassic.h b/include/linux/rcuclassic.h index f3f697df1d71..80044a4f3ab9 100644 --- a/include/linux/rcuclassic.h +++ b/include/linux/rcuclassic.h @@ -181,4 +181,10 @@ extern long rcu_batches_completed_bh(void); #define rcu_enter_nohz() do { } while (0) #define rcu_exit_nohz() do { } while (0) +/* A context switch is a grace period for rcuclassic. */ +static inline int rcu_blocking_is_gp(void) +{ + return num_online_cpus() == 1; +} + #endif /* __LINUX_RCUCLASSIC_H */ diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h index 921340a7b71c..528343e6da51 100644 --- a/include/linux/rcupdate.h +++ b/include/linux/rcupdate.h @@ -52,6 +52,9 @@ struct rcu_head { void (*func)(struct rcu_head *head); }; +/* Internal to kernel, but needed by rcupreempt.h. */ +extern int rcu_scheduler_active; + #if defined(CONFIG_CLASSIC_RCU) #include #elif defined(CONFIG_TREE_RCU) @@ -265,6 +268,7 @@ extern void rcu_barrier_sched(void); /* Internal to kernel */ extern void rcu_init(void); +extern void rcu_scheduler_starting(void); extern int rcu_needs_cpu(int cpu); #endif /* __LINUX_RCUPDATE_H */ diff --git a/include/linux/rcupreempt.h b/include/linux/rcupreempt.h index 3e05c09b54a2..74304b4538d8 100644 --- a/include/linux/rcupreempt.h +++ b/include/linux/rcupreempt.h @@ -142,4 +142,19 @@ static inline void rcu_exit_nohz(void) #define rcu_exit_nohz() do { } while (0) #endif /* CONFIG_NO_HZ */ +/* + * A context switch is a grace period for rcupreempt synchronize_rcu() + * only during early boot, before the scheduler has been initialized. + * So, how the heck do we get a context switch? Well, if the caller + * invokes synchronize_rcu(), they are willing to accept a context + * switch, so we simply pretend that one happened. + * + * After boot, there might be a blocked or preempted task in an RCU + * read-side critical section, so we cannot then take the fastpath. + */ +static inline int rcu_blocking_is_gp(void) +{ + return num_online_cpus() == 1 && !rcu_scheduler_active; +} + #endif /* __LINUX_RCUPREEMPT_H */ diff --git a/include/linux/rcutree.h b/include/linux/rcutree.h index d4368b7975c3..a722fb67bb2d 100644 --- a/include/linux/rcutree.h +++ b/include/linux/rcutree.h @@ -326,4 +326,10 @@ static inline void rcu_exit_nohz(void) } #endif /* CONFIG_NO_HZ */ +/* A context switch is a grace period for rcutree. */ +static inline int rcu_blocking_is_gp(void) +{ + return num_online_cpus() == 1; +} + #endif /* __LINUX_RCUTREE_H */ diff --git a/init/main.c b/init/main.c index 844209453c02..83697e160b3a 100644 --- a/init/main.c +++ b/init/main.c @@ -97,7 +97,7 @@ static inline void mark_rodata_ro(void) { } extern void tc_init(void); #endif -enum system_states system_state; +enum system_states system_state __read_mostly; EXPORT_SYMBOL(system_state); /* @@ -463,6 +463,7 @@ static noinline void __init_refok rest_init(void) * at least once to get things moving: */ init_idle_bootup_task(current); + rcu_scheduler_starting(); preempt_enable_no_resched(); schedule(); preempt_disable(); diff --git a/kernel/rcuclassic.c b/kernel/rcuclassic.c index bd5a9003497c..654c640a6b9c 100644 --- a/kernel/rcuclassic.c +++ b/kernel/rcuclassic.c @@ -679,8 +679,8 @@ int rcu_needs_cpu(int cpu) void rcu_check_callbacks(int cpu, int user) { if (user || - (idle_cpu(cpu) && !in_softirq() && - hardirq_count() <= (1 << HARDIRQ_SHIFT))) { + (idle_cpu(cpu) && rcu_scheduler_active && + !in_softirq() && hardirq_count() <= (1 << HARDIRQ_SHIFT))) { /* * Get here if this CPU took its interrupt from user diff --git a/kernel/rcupdate.c b/kernel/rcupdate.c index d92a76a881aa..cae8a059cf47 100644 --- a/kernel/rcupdate.c +++ b/kernel/rcupdate.c @@ -44,6 +44,7 @@ #include #include #include +#include enum rcu_barrier { RCU_BARRIER_STD, @@ -55,6 +56,7 @@ static DEFINE_PER_CPU(struct rcu_head, rcu_barrier_head) = {NULL}; static atomic_t rcu_barrier_cpu_count; static DEFINE_MUTEX(rcu_barrier_mutex); static struct completion rcu_barrier_completion; +int rcu_scheduler_active __read_mostly; /* * Awaken the corresponding synchronize_rcu() instance now that a @@ -80,6 +82,10 @@ void wakeme_after_rcu(struct rcu_head *head) void synchronize_rcu(void) { struct rcu_synchronize rcu; + + if (rcu_blocking_is_gp()) + return; + init_completion(&rcu.completion); /* Will wake me after RCU finished. */ call_rcu(&rcu.head, wakeme_after_rcu); @@ -175,3 +181,9 @@ void __init rcu_init(void) __rcu_init(); } +void rcu_scheduler_starting(void) +{ + WARN_ON(num_online_cpus() != 1); + WARN_ON(nr_context_switches() > 0); + rcu_scheduler_active = 1; +} diff --git a/kernel/rcupreempt.c b/kernel/rcupreempt.c index 33cfc50781f9..5d59e850fb71 100644 --- a/kernel/rcupreempt.c +++ b/kernel/rcupreempt.c @@ -1181,6 +1181,9 @@ void __synchronize_sched(void) { struct rcu_synchronize rcu; + if (num_online_cpus() == 1) + return; /* blocking is gp if only one CPU! */ + init_completion(&rcu.completion); /* Will wake me after RCU finished. */ call_rcu_sched(&rcu.head, wakeme_after_rcu); diff --git a/kernel/rcutree.c b/kernel/rcutree.c index b2fd602a6f6f..97ce31579ec0 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -948,8 +948,8 @@ static void rcu_do_batch(struct rcu_data *rdp) void rcu_check_callbacks(int cpu, int user) { if (user || - (idle_cpu(cpu) && !in_softirq() && - hardirq_count() <= (1 << HARDIRQ_SHIFT))) { + (idle_cpu(cpu) && rcu_scheduler_active && + !in_softirq() && hardirq_count() <= (1 << HARDIRQ_SHIFT))) { /* * Get here if this CPU took its interrupt from user From 8b1a8ff8b321a9384304aeea4dbdb9747daf7ee8 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 28 Feb 2009 00:08:53 -0500 Subject: [PATCH 27/92] ext4: Remove duplicate call to ext4_commit_super() in ext4_freeze() Commit c4be0c1d added error checking to ext4_freeze() when calling ext4_commit_super(). Unfortunately the patch failed to remove the original call to ext4_commit_super(), with the net result that when freezing the filesystem, the superblock gets written twice, the first time without error checking. Signed-off-by: "Theodore Ts'o" --- fs/ext4/super.c | 1 - 1 file changed, 1 deletion(-) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index a5732c58f676..39d1993cfa13 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -3091,7 +3091,6 @@ static int ext4_freeze(struct super_block *sb) /* Journal blocked and flushed, clear needs_recovery flag. */ EXT4_CLEAR_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER); - ext4_commit_super(sb, EXT4_SB(sb)->s_es, 1); error = ext4_commit_super(sb, EXT4_SB(sb)->s_es, 1); if (error) goto out; From d8ae4601a4b7ea1fa17fa395c3468c0e144d1275 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 28 Feb 2009 09:50:01 -0500 Subject: [PATCH 28/92] ext4: Reorder fs/Makefile so that ext2 root fs's are mounted using ext2 In fs/Makefile, ext3 was placed before ext2 so that a root filesystem that possessed a journal, it would be mounted as ext3 instead of ext2. This was necessary because a cleanly unmounted ext3 filesystem was fully backwards compatible with ext2, and could be mounted by ext2 --- but it was desirable that it be mounted with ext3 so that the journaling would be enabled. The ext4 filesystem supports new incompatible features, so there is no danger of an ext4 filesystem being mistaken for an ext2 filesystem. At that point, the relative ordering of ext4 with respect to ext2 didn't matter until ext4 gained the ability to mount filesystems without a journal starting in 2.6.29-rc1. Now that this is the case, given that ext4 is before ext2, it means that root filesystems that were using the plain-jane ext2 format are getting mounted using the ext4 filesystem driver, which is a change in behavior which could be surprising to users. It's doubtful that there are that many ext2-only root filesystem users that would also have ext4 compiled into the kernel, but to adhere to the principle of least surprise, the correct ordering in fs/Makefile is ext3, followed by ext2, and finally ext4. Signed-off-by: "Theodore Ts'o" --- fs/Makefile | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/fs/Makefile b/fs/Makefile index 38bc735c67ad..dc20db348679 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -69,10 +69,12 @@ obj-$(CONFIG_DLM) += dlm/ # Do not add any filesystems before this line obj-$(CONFIG_REISERFS_FS) += reiserfs/ obj-$(CONFIG_EXT3_FS) += ext3/ # Before ext2 so root fs can be ext3 -obj-$(CONFIG_EXT4_FS) += ext4/ # Before ext2 so root fs can be ext4 +obj-$(CONFIG_EXT2_FS) += ext2/ +# We place ext4 after ext2 so plain ext2 root fs's are mounted using ext2 +# unless explicitly requested by rootfstype +obj-$(CONFIG_EXT4_FS) += ext4/ obj-$(CONFIG_JBD) += jbd/ obj-$(CONFIG_JBD2) += jbd2/ -obj-$(CONFIG_EXT2_FS) += ext2/ obj-$(CONFIG_CRAMFS) += cramfs/ obj-$(CONFIG_SQUASHFS) += squashfs/ obj-y += ramfs/ From 8f64b32eb73fbfe9f38c4123121b63ee409278a7 Mon Sep 17 00:00:00 2001 From: Eric Sandeen Date: Thu, 26 Feb 2009 00:57:35 -0500 Subject: [PATCH 29/92] ext4: don't call jbd2_journal_force_commit_nested without journal Running without a journal, I oopsed when I ran out of space, because we called jbd2_journal_force_commit_nested() from ext4_should_retry_alloc() without a journal. This should take care of it, I think. Signed-off-by: Eric Sandeen Signed-off-by: "Theodore Ts'o" --- fs/ext4/balloc.c | 4 +++- fs/ext4/inode.c | 2 +- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 9a50b8052dcf..de9459b4cb94 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -609,7 +609,9 @@ int ext4_claim_free_blocks(struct ext4_sb_info *sbi, */ int ext4_should_retry_alloc(struct super_block *sb, int *retries) { - if (!ext4_has_free_blocks(EXT4_SB(sb), 1) || (*retries)++ > 3) + if (!ext4_has_free_blocks(EXT4_SB(sb), 1) || + (*retries)++ > 3 || + !EXT4_SB(sb)->s_journal) return 0; jbd_debug(1, "%s: retrying operation after ENOSPC\n", sb->s_id); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 51cdd13e1c31..c7fed5b18745 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -2544,7 +2544,7 @@ retry: ext4_journal_stop(handle); - if (mpd.retval == -ENOSPC) { + if ((mpd.retval == -ENOSPC) && sbi->s_journal) { /* commit the transaction which would * free blocks released in the transaction * and try again From a760a6656e6f00bb0144a42a048cf0266646e22c Mon Sep 17 00:00:00 2001 From: Herbert Xu Date: Thu, 26 Feb 2009 14:06:31 +0800 Subject: [PATCH 30/92] crypto: api - Fix module load deadlock with fallback algorithms With the mandatory algorithm testing at registration, we have now created a deadlock with algorithms requiring fallbacks. This can happen if the module containing the algorithm requiring fallback is loaded first, without the fallback module being loaded first. The system will then try to test the new algorithm, find that it needs to load a fallback, and then try to load that. As both algorithms share the same module alias, it can attempt to load the original algorithm again and block indefinitely. As algorithms requiring fallbacks are a special case, we can fix this by giving them a different module alias than the rest. Then it's just a matter of using the right aliases according to what algorithms we're trying to find. Signed-off-by: Herbert Xu --- arch/s390/crypto/aes_s390.c | 2 +- crypto/api.c | 15 +++++++++++++-- drivers/crypto/padlock-aes.c | 2 +- drivers/crypto/padlock-sha.c | 4 ++-- 4 files changed, 17 insertions(+), 6 deletions(-) diff --git a/arch/s390/crypto/aes_s390.c b/arch/s390/crypto/aes_s390.c index c42cd898f68b..6118890c946d 100644 --- a/arch/s390/crypto/aes_s390.c +++ b/arch/s390/crypto/aes_s390.c @@ -556,7 +556,7 @@ static void __exit aes_s390_fini(void) module_init(aes_s390_init); module_exit(aes_s390_fini); -MODULE_ALIAS("aes"); +MODULE_ALIAS("aes-all"); MODULE_DESCRIPTION("Rijndael (AES) Cipher Algorithm"); MODULE_LICENSE("GPL"); diff --git a/crypto/api.c b/crypto/api.c index efe77df6863f..38a2bc02a98c 100644 --- a/crypto/api.c +++ b/crypto/api.c @@ -215,8 +215,19 @@ struct crypto_alg *crypto_larval_lookup(const char *name, u32 type, u32 mask) mask &= ~(CRYPTO_ALG_LARVAL | CRYPTO_ALG_DEAD); type &= mask; - alg = try_then_request_module(crypto_alg_lookup(name, type, mask), - name); + alg = crypto_alg_lookup(name, type, mask); + if (!alg) { + char tmp[CRYPTO_MAX_ALG_NAME]; + + request_module(name); + + if (!((type ^ CRYPTO_ALG_NEED_FALLBACK) & mask) && + snprintf(tmp, sizeof(tmp), "%s-all", name) < sizeof(tmp)) + request_module(tmp); + + alg = crypto_alg_lookup(name, type, mask); + } + if (alg) return crypto_is_larval(alg) ? crypto_larval_wait(alg) : alg; diff --git a/drivers/crypto/padlock-aes.c b/drivers/crypto/padlock-aes.c index 856b3cc25583..3f0fdd18255d 100644 --- a/drivers/crypto/padlock-aes.c +++ b/drivers/crypto/padlock-aes.c @@ -489,4 +489,4 @@ MODULE_DESCRIPTION("VIA PadLock AES algorithm support"); MODULE_LICENSE("GPL"); MODULE_AUTHOR("Michal Ludvig"); -MODULE_ALIAS("aes"); +MODULE_ALIAS("aes-all"); diff --git a/drivers/crypto/padlock-sha.c b/drivers/crypto/padlock-sha.c index a7fbadebf623..a2c8e8514b63 100644 --- a/drivers/crypto/padlock-sha.c +++ b/drivers/crypto/padlock-sha.c @@ -304,7 +304,7 @@ MODULE_DESCRIPTION("VIA PadLock SHA1/SHA256 algorithms support."); MODULE_LICENSE("GPL"); MODULE_AUTHOR("Michal Ludvig"); -MODULE_ALIAS("sha1"); -MODULE_ALIAS("sha256"); +MODULE_ALIAS("sha1-all"); +MODULE_ALIAS("sha256-all"); MODULE_ALIAS("sha1-padlock"); MODULE_ALIAS("sha256-padlock"); From cac64d00c256e65776d575e82aaf540632b66178 Mon Sep 17 00:00:00 2001 From: Hiroshi Shimamoto Date: Wed, 25 Feb 2009 09:59:26 -0800 Subject: [PATCH 31/92] sched_rt: don't start timer when rt bandwidth disabled Impact: fix incorrect condition check No need to start rt bandwidth timer when rt bandwidth is disabled. If this timer starts, it may stop at sched_rt_period_timer() on the first time. Signed-off-by: Hiroshi Shimamoto Acked-by: Peter Zijlstra Signed-off-by: Ingo Molnar --- kernel/sched.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/kernel/sched.c b/kernel/sched.c index 410eec404133..c3baa9653d1d 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -223,7 +223,7 @@ static void start_rt_bandwidth(struct rt_bandwidth *rt_b) { ktime_t now; - if (rt_bandwidth_enabled() && rt_b->rt_runtime == RUNTIME_INF) + if (!rt_bandwidth_enabled() || rt_b->rt_runtime == RUNTIME_INF) return; if (hrtimer_active(&rt_b->rt_period_timer)) From 54e991242850edc8c53f71fa5aa3ba7a93ce38f5 Mon Sep 17 00:00:00 2001 From: Dhaval Giani Date: Fri, 27 Feb 2009 15:13:54 +0530 Subject: [PATCH 32/92] sched: don't allow setuid to succeed if the user does not have rt bandwidth Impact: fix hung task with certain (non-default) rt-limit settings Corey Hickey reported that on using setuid to change the uid of a rt process, the process would be unkillable and not be running. This is because there was no rt runtime for that user group. Add in a check to see if a user can attach an rt task to its task group. On failure, return EINVAL, which is also returned in CONFIG_CGROUP_SCHED. Reported-by: Corey Hickey Signed-off-by: Dhaval Giani Acked-by: Peter Zijlstra Signed-off-by: Ingo Molnar --- include/linux/sched.h | 4 ++++ kernel/sched.c | 13 +++++++++++-- kernel/sys.c | 31 ++++++++++++++++++++----------- kernel/user.c | 18 ++++++++++++++++++ 4 files changed, 53 insertions(+), 13 deletions(-) diff --git a/include/linux/sched.h b/include/linux/sched.h index 8981e52c714f..8c216e057c94 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -2291,9 +2291,13 @@ extern long sched_group_rt_runtime(struct task_group *tg); extern int sched_group_set_rt_period(struct task_group *tg, long rt_period_us); extern long sched_group_rt_period(struct task_group *tg); +extern int sched_rt_can_attach(struct task_group *tg, struct task_struct *tsk); #endif #endif +extern int task_can_switch_user(struct user_struct *up, + struct task_struct *tsk); + #ifdef CONFIG_TASK_XACCT static inline void add_rchar(struct task_struct *tsk, ssize_t amt) { diff --git a/kernel/sched.c b/kernel/sched.c index c3baa9653d1d..8e2558c2ba67 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -9224,6 +9224,16 @@ static int sched_rt_global_constraints(void) return ret; } + +int sched_rt_can_attach(struct task_group *tg, struct task_struct *tsk) +{ + /* Don't accept realtime tasks when there is no way for them to run */ + if (rt_task(tsk) && tg->rt_bandwidth.rt_runtime == 0) + return 0; + + return 1; +} + #else /* !CONFIG_RT_GROUP_SCHED */ static int sched_rt_global_constraints(void) { @@ -9317,8 +9327,7 @@ cpu_cgroup_can_attach(struct cgroup_subsys *ss, struct cgroup *cgrp, struct task_struct *tsk) { #ifdef CONFIG_RT_GROUP_SCHED - /* Don't accept realtime tasks when there is no way for them to run */ - if (rt_task(tsk) && cgroup_tg(cgrp)->rt_bandwidth.rt_runtime == 0) + if (!sched_rt_can_attach(cgroup_tg(cgrp), tsk)) return -EINVAL; #else /* We don't support RT-tasks being in separate groups */ diff --git a/kernel/sys.c b/kernel/sys.c index f145c415bc16..37f458e6882a 100644 --- a/kernel/sys.c +++ b/kernel/sys.c @@ -559,7 +559,7 @@ error: abort_creds(new); return retval; } - + /* * change the user struct in a credentials set to match the new UID */ @@ -571,6 +571,11 @@ static int set_user(struct cred *new) if (!new_user) return -EAGAIN; + if (!task_can_switch_user(new_user, current)) { + free_uid(new_user); + return -EINVAL; + } + if (atomic_read(&new_user->processes) >= current->signal->rlim[RLIMIT_NPROC].rlim_cur && new_user != INIT_USER) { @@ -631,10 +636,11 @@ SYSCALL_DEFINE2(setreuid, uid_t, ruid, uid_t, euid) goto error; } - retval = -EAGAIN; - if (new->uid != old->uid && set_user(new) < 0) - goto error; - + if (new->uid != old->uid) { + retval = set_user(new); + if (retval < 0) + goto error; + } if (ruid != (uid_t) -1 || (euid != (uid_t) -1 && euid != old->uid)) new->suid = new->euid; @@ -680,9 +686,10 @@ SYSCALL_DEFINE1(setuid, uid_t, uid) retval = -EPERM; if (capable(CAP_SETUID)) { new->suid = new->uid = uid; - if (uid != old->uid && set_user(new) < 0) { - retval = -EAGAIN; - goto error; + if (uid != old->uid) { + retval = set_user(new); + if (retval < 0) + goto error; } } else if (uid != old->uid && uid != new->suid) { goto error; @@ -734,11 +741,13 @@ SYSCALL_DEFINE3(setresuid, uid_t, ruid, uid_t, euid, uid_t, suid) goto error; } - retval = -EAGAIN; if (ruid != (uid_t) -1) { new->uid = ruid; - if (ruid != old->uid && set_user(new) < 0) - goto error; + if (ruid != old->uid) { + retval = set_user(new); + if (retval < 0) + goto error; + } } if (euid != (uid_t) -1) new->euid = euid; diff --git a/kernel/user.c b/kernel/user.c index 3551ac742395..6a9b696128c8 100644 --- a/kernel/user.c +++ b/kernel/user.c @@ -362,6 +362,24 @@ static void free_user(struct user_struct *up, unsigned long flags) #endif +#if defined(CONFIG_RT_GROUP_SCHED) && defined(CONFIG_USER_SCHED) +/* + * We need to check if a setuid can take place. This function should be called + * before successfully completing the setuid. + */ +int task_can_switch_user(struct user_struct *up, struct task_struct *tsk) +{ + + return sched_rt_can_attach(up->tg, tsk); + +} +#else +int task_can_switch_user(struct user_struct *up, struct task_struct *tsk) +{ + return 1; +} +#endif + /* * Locate the user_struct for the passed UID. If found, take a ref on it. The * caller must undo that ref with free_uid(). From ea18aa464452c3e6550320d247c0306aaa2d156f Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Fri, 27 Feb 2009 17:36:33 +0100 Subject: [PATCH 33/92] ALSA: hda - Fix digital mic on dell-m4-1 and dell-m4-3 Fix num_dmuxes initialization for dell-m4-1 and dell-m4-3 models of IDT 92HD71bxx codec, which was wrongly set to zero. Cc: Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_sigmatel.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/pci/hda/patch_sigmatel.c b/sound/pci/hda/patch_sigmatel.c index 8027edf3c8f2..3bc427645da8 100644 --- a/sound/pci/hda/patch_sigmatel.c +++ b/sound/pci/hda/patch_sigmatel.c @@ -4989,7 +4989,7 @@ again: case STAC_DELL_M4_3: spec->num_dmics = 1; spec->num_smuxes = 0; - spec->num_dmuxes = 0; + spec->num_dmuxes = 1; break; default: spec->num_dmics = STAC92HD71BXX_NUM_DMICS; From bb543c969467f33c3a1a0ccfcfcd9a508cd81c54 Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Fri, 27 Feb 2009 17:44:07 +0100 Subject: [PATCH 34/92] ALSA: hda - Add quirk for new HP xw series Added model=hp-bpc for new HP xw series (103c:170b). Signed-off-by: Takashi Iwai --- sound/pci/hda/patch_realtek.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/pci/hda/patch_realtek.c b/sound/pci/hda/patch_realtek.c index a680be0d4534..6c26afcb8262 100644 --- a/sound/pci/hda/patch_realtek.c +++ b/sound/pci/hda/patch_realtek.c @@ -10557,6 +10557,7 @@ static struct snd_pci_quirk alc262_cfg_tbl[] = { SND_PCI_QUIRK(0x103c, 0x1309, "HP xw4*00", ALC262_HP_BPC), SND_PCI_QUIRK(0x103c, 0x130a, "HP xw6*00", ALC262_HP_BPC), SND_PCI_QUIRK(0x103c, 0x130b, "HP xw8*00", ALC262_HP_BPC), + SND_PCI_QUIRK(0x103c, 0x170b, "HP xw*", ALC262_HP_BPC), SND_PCI_QUIRK(0x103c, 0x2800, "HP D7000", ALC262_HP_BPC_D7000_WL), SND_PCI_QUIRK(0x103c, 0x2801, "HP D7000", ALC262_HP_BPC_D7000_WF), SND_PCI_QUIRK(0x103c, 0x2802, "HP D7000", ALC262_HP_BPC_D7000_WL), From 92b9af9e4f144535c65aee673cfad309f25fa465 Mon Sep 17 00:00:00 2001 From: Ingo Molnar Date: Sat, 28 Feb 2009 14:09:27 +0100 Subject: [PATCH 35/92] x86: i915 needs pgprot_writecombine() and is_io_mapping_possible() Impact: build fix Theodore Ts reported that the i915 driver needs these symbols: ERROR: "pgprot_writecombine" [drivers/gpu/drm/i915/i915.ko] undefined! ERROR: "is_io_mapping_possible" [drivers/gpu/drm/i915/i915.ko] undefined! Reported-by: Theodore Ts'o wrote: Signed-off-by: Ingo Molnar --- arch/x86/mm/iomap_32.c | 15 ++++----------- arch/x86/mm/pat.c | 2 ++ 2 files changed, 6 insertions(+), 11 deletions(-) diff --git a/arch/x86/mm/iomap_32.c b/arch/x86/mm/iomap_32.c index 6c2b1af16926..04102d42ff42 100644 --- a/arch/x86/mm/iomap_32.c +++ b/arch/x86/mm/iomap_32.c @@ -20,23 +20,16 @@ #include #include -#ifdef CONFIG_X86_PAE -int -is_io_mapping_possible(resource_size_t base, unsigned long size) -{ - return 1; -} -#else -int -is_io_mapping_possible(resource_size_t base, unsigned long size) +int is_io_mapping_possible(resource_size_t base, unsigned long size) { +#ifndef CONFIG_X86_PAE /* There is no way to map greater than 1 << 32 address without PAE */ if (base + size > 0x100000000ULL) return 0; - +#endif return 1; } -#endif +EXPORT_SYMBOL_GPL(is_io_mapping_possible); /* Map 'pfn' using fixed map 'type' and protections 'prot' */ diff --git a/arch/x86/mm/pat.c b/arch/x86/mm/pat.c index aebbf67a79d0..e0ab173b6974 100644 --- a/arch/x86/mm/pat.c +++ b/arch/x86/mm/pat.c @@ -11,6 +11,7 @@ #include #include #include +#include #include #include #include @@ -868,6 +869,7 @@ pgprot_t pgprot_writecombine(pgprot_t prot) else return pgprot_noncached(prot); } +EXPORT_SYMBOL_GPL(pgprot_writecombine); #if defined(CONFIG_DEBUG_FS) && defined(CONFIG_X86_PAT) From 139ebe8dc80dd74cb2ac9f5603d18fbf5cff049f Mon Sep 17 00:00:00 2001 From: Daniel Ritz Date: Sat, 28 Feb 2009 12:50:54 -0800 Subject: [PATCH 36/92] Input: usbtouchscreen - fix eGalax HID ignoring Commit ec42d4481e36cbdb5b2801f957e678211a9e5ae2 broke usbtouchscreen for some eGalax/EETI devices that claim to be HID, but are not. Devices confirmed to be real HID have the class set to HID and the protocol set to 'mouse'. Some have HID class but protocol set to 'none'. Those are not HID and should be driven by usbtouchscreen. Fix the device ignoring macro by adding match for the protocol too. Signed-off-by: Daniel Ritz Signed-off-by: Dmitry Torokhov --- drivers/input/touchscreen/usbtouchscreen.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/input/touchscreen/usbtouchscreen.c b/drivers/input/touchscreen/usbtouchscreen.c index 6d27a1d661e6..fb7cb9bdfbd5 100644 --- a/drivers/input/touchscreen/usbtouchscreen.c +++ b/drivers/input/touchscreen/usbtouchscreen.c @@ -122,6 +122,7 @@ enum { #define USB_DEVICE_HID_CLASS(vend, prod) \ .match_flags = USB_DEVICE_ID_MATCH_INT_CLASS \ + | USB_DEVICE_ID_MATCH_INT_PROTOCOL \ | USB_DEVICE_ID_MATCH_DEVICE, \ .idVendor = (vend), \ .idProduct = (prod), \ From 4d368456808c977b8e9782dbe9542cf8ddedbab8 Mon Sep 17 00:00:00 2001 From: Andy Whitcroft Date: Sat, 28 Feb 2009 12:51:01 -0800 Subject: [PATCH 37/92] Input: synaptics - ensure we reset the device on resume When resuming from suspend newer Synaptics touchpads do not recover correctly. Analysis of the resume sequence as applied in Linux was compared to that of other operating systems. This indicated that the other OSs were resetting the mouse before attempting to detect it (for all Synaptics touchpads, old and new). Applying this same modification fixes these newer Synaptics touchpads and brings the driver into line with common OS reset behaviour. Signed-off-by: Andy Whitcroft Signed-off-by: Andrew Morton Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/synaptics.c | 9 ++++----- 1 file changed, 4 insertions(+), 5 deletions(-) diff --git a/drivers/input/mouse/synaptics.c b/drivers/input/mouse/synaptics.c index 865fc69e9bc3..f3e4f7b0240d 100644 --- a/drivers/input/mouse/synaptics.c +++ b/drivers/input/mouse/synaptics.c @@ -182,11 +182,6 @@ static int synaptics_identify(struct psmouse *psmouse) static int synaptics_query_hardware(struct psmouse *psmouse) { - int retries = 0; - - while ((retries++ < 3) && psmouse_reset(psmouse)) - /* empty */; - if (synaptics_identify(psmouse)) return -1; if (synaptics_model_id(psmouse)) @@ -582,6 +577,8 @@ static int synaptics_reconnect(struct psmouse *psmouse) struct synaptics_data *priv = psmouse->private; struct synaptics_data old_priv = *priv; + psmouse_reset(psmouse); + if (synaptics_detect(psmouse, 0)) return -1; @@ -640,6 +637,8 @@ int synaptics_init(struct psmouse *psmouse) if (!priv) return -1; + psmouse_reset(psmouse); + if (synaptics_query_hardware(psmouse)) { printk(KERN_ERR "Unable to query Synaptics hardware.\n"); goto init_fail; From 9ab7b25e6a30d2292bd6d4913b71c918ee1e21b4 Mon Sep 17 00:00:00 2001 From: Arjan Opmeer Date: Sat, 28 Feb 2009 13:52:40 -0800 Subject: [PATCH 38/92] Input: elantech - touchpad driver miss-recognising logitech mice Some Logitech mice react to the magic knock like Elantech touchpad would. This leads to those mice being misdetected as Elantech touchpads. Add a version query to elantech_detect() to distinguish the two. [dtor@mail.ru: - lower severity of some messages - when we are not sure yet if device is Elantech or not not responding to knock is not an error. ] Signed-off-by: Arjan Opmeer Signed-off-by: Dmitry Torokhov --- drivers/input/mouse/elantech.c | 32 ++++++++++++++++++++++++-------- 1 file changed, 24 insertions(+), 8 deletions(-) diff --git a/drivers/input/mouse/elantech.c b/drivers/input/mouse/elantech.c index b9a25d57bc5e..6ab0eb1ada1c 100644 --- a/drivers/input/mouse/elantech.c +++ b/drivers/input/mouse/elantech.c @@ -542,7 +542,7 @@ int elantech_detect(struct psmouse *psmouse, int set_properties) ps2_command(ps2dev, NULL, PSMOUSE_CMD_SETSCALE11) || ps2_command(ps2dev, NULL, PSMOUSE_CMD_SETSCALE11) || ps2_command(ps2dev, param, PSMOUSE_CMD_GETINFO)) { - pr_err("elantech.c: sending Elantech magic knock failed.\n"); + pr_debug("elantech.c: sending Elantech magic knock failed.\n"); return -1; } @@ -551,8 +551,27 @@ int elantech_detect(struct psmouse *psmouse, int set_properties) * set of magic numbers */ if (param[0] != 0x3c || param[1] != 0x03 || param[2] != 0xc8) { - pr_info("elantech.c: unexpected magic knock result 0x%02x, 0x%02x, 0x%02x.\n", - param[0], param[1], param[2]); + pr_debug("elantech.c: " + "unexpected magic knock result 0x%02x, 0x%02x, 0x%02x.\n", + param[0], param[1], param[2]); + return -1; + } + + /* + * Query touchpad's firmware version and see if it reports known + * value to avoid mis-detection. Logitech mice are known to respond + * to Elantech magic knock and there might be more. + */ + if (synaptics_send_cmd(psmouse, ETP_FW_VERSION_QUERY, param)) { + pr_debug("elantech.c: failed to query firmware version.\n"); + return -1; + } + + pr_debug("elantech.c: Elantech version query result 0x%02x, 0x%02x, 0x%02x.\n", + param[0], param[1], param[2]); + + if (param[0] == 0 || param[1] != 0) { + pr_debug("elantech.c: Probably not a real Elantech touchpad. Aborting.\n"); return -1; } @@ -600,8 +619,7 @@ int elantech_init(struct psmouse *psmouse) int i, error; unsigned char param[3]; - etd = kzalloc(sizeof(struct elantech_data), GFP_KERNEL); - psmouse->private = etd; + psmouse->private = etd = kzalloc(sizeof(struct elantech_data), GFP_KERNEL); if (!etd) return -1; @@ -610,14 +628,12 @@ int elantech_init(struct psmouse *psmouse) etd->parity[i] = etd->parity[i & (i - 1)] ^ 1; /* - * Find out what version hardware this is + * Do the version query again so we can store the result */ if (synaptics_send_cmd(psmouse, ETP_FW_VERSION_QUERY, param)) { pr_err("elantech.c: failed to query firmware version.\n"); goto init_fail; } - pr_info("elantech.c: Elantech version query result 0x%02x, 0x%02x, 0x%02x.\n", - param[0], param[1], param[2]); etd->fw_version_maj = param[0]; etd->fw_version_min = param[2]; From d7f59dc4642ce2fc7b79fcd4ec02ffce7f21eb02 Mon Sep 17 00:00:00 2001 From: Paul Moore Date: Fri, 27 Feb 2009 15:00:03 -0500 Subject: [PATCH 39/92] selinux: Fix a panic in selinux_netlbl_inode_permission() Rick McNeal from LSI identified a panic in selinux_netlbl_inode_permission() caused by a certain sequence of SUNRPC operations. The problem appears to be due to the lack of NULL pointer checking in the function; this patch adds the pointer checks so the function will exit safely in the cases where the socket is not completely initialized. Signed-off-by: Paul Moore Signed-off-by: James Morris --- security/selinux/netlabel.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/security/selinux/netlabel.c b/security/selinux/netlabel.c index 3f4b26647386..350794ab9b42 100644 --- a/security/selinux/netlabel.c +++ b/security/selinux/netlabel.c @@ -386,11 +386,12 @@ int selinux_netlbl_inode_permission(struct inode *inode, int mask) if (!S_ISSOCK(inode->i_mode) || ((mask & (MAY_WRITE | MAY_APPEND)) == 0)) return 0; - sock = SOCKET_I(inode); sk = sock->sk; + if (sk == NULL) + return 0; sksec = sk->sk_security; - if (sksec->nlbl_state != NLBL_REQUIRE) + if (sksec == NULL || sksec->nlbl_state != NLBL_REQUIRE) return 0; local_bh_disable(); From 38f1df27e3191d76e983cb9c6b4392582fd32fda Mon Sep 17 00:00:00 2001 From: Takashi Iwai Date: Sun, 1 Mar 2009 10:55:44 +0100 Subject: [PATCH 40/92] ALSA: hda - Add probe_mask default for Toshiba laptop with ALC268 Some Toshiba laptops have another ALC268 codec on slot#3 that conflicts with the primary codec. The codec#3 is for the digital I/O, and should be fixed by the driver, but it'd need a bunch of changes. So, let's fix the probe problem temporarily by setting the default probe_mask value. Reference: kernel bugzilla #12735 http://bugzilla.kernel.org/show_bug.cgi?id=12735 Tested-by: Alexey Dobriyan Signed-off-by: Takashi Iwai --- sound/pci/hda/hda_intel.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c index c8d9178f47e5..5e909e0da04b 100644 --- a/sound/pci/hda/hda_intel.c +++ b/sound/pci/hda/hda_intel.c @@ -2095,6 +2095,8 @@ static struct snd_pci_quirk probe_mask_list[] __devinitdata = { SND_PCI_QUIRK(0x1028, 0x20ac, "Dell Studio Desktop", 0x01), /* including bogus ALC268 in slot#2 that conflicts with ALC888 */ SND_PCI_QUIRK(0x17c0, 0x4085, "Medion MD96630", 0x01), + /* conflict of ALC268 in slot#3 (digital I/O); a temporary fix */ + SND_PCI_QUIRK(0x1179, 0xff00, "Toshiba laptop", 0x03), {} }; From 5ce04e3de8c36ba37c56e94e3c4dc7973c7f546c Mon Sep 17 00:00:00 2001 From: "Pallipadi, Venkatesh" Date: Sun, 1 Mar 2009 08:53:27 -0800 Subject: [PATCH 41/92] fix warning in io_mapping_map_wc() Signed-off-by: Venkatesh Pallipadi Signed-off-by: Ingo Molnar --- include/linux/io-mapping.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/include/linux/io-mapping.h b/include/linux/io-mapping.h index cbc2f0cd631b..0adb0f91568c 100644 --- a/include/linux/io-mapping.h +++ b/include/linux/io-mapping.h @@ -91,8 +91,11 @@ io_mapping_unmap_atomic(void *vaddr) static inline void * io_mapping_map_wc(struct io_mapping *mapping, unsigned long offset) { + resource_size_t phys_addr; + BUG_ON(offset >= mapping->size); - resource_size_t phys_addr = mapping->base + offset; + phys_addr = mapping->base + offset; + return ioremap_wc(phys_addr, PAGE_SIZE); } From fab852aaf761a00cfe16330429b7cac15cceaeb9 Mon Sep 17 00:00:00 2001 From: Pekka Paalanen Date: Sun, 1 Mar 2009 16:09:14 +0200 Subject: [PATCH 42/92] x86: count errors in testmmiotrace.ko Check the read values against the written values in the MMIO read/write test. This test shows if the given MMIO test area really works as memory, which is a prerequisite for a successful mmiotrace test. Signed-off-by: Pekka Paalanen Cc: Stuart Bennett Cc: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/mm/testmmiotrace.c | 35 +++++++++++++++++++++++++++++------ 1 file changed, 29 insertions(+), 6 deletions(-) diff --git a/arch/x86/mm/testmmiotrace.c b/arch/x86/mm/testmmiotrace.c index ab50a8d7402c..4b29f3b0ee01 100644 --- a/arch/x86/mm/testmmiotrace.c +++ b/arch/x86/mm/testmmiotrace.c @@ -1,5 +1,5 @@ /* - * Written by Pekka Paalanen, 2008 + * Written by Pekka Paalanen, 2008-2009 */ #include #include @@ -11,28 +11,51 @@ static unsigned long mmio_address; module_param(mmio_address, ulong, 0); MODULE_PARM_DESC(mmio_address, "Start address of the mapping of 16 kB."); +static unsigned v16(unsigned i) +{ + return i * 12 + 7; +} + +static unsigned v32(unsigned i) +{ + return i * 212371 + 13; +} + static void do_write_test(void __iomem *p) { unsigned int i; mmiotrace_printk("Write test.\n"); + for (i = 0; i < 256; i++) iowrite8(i, p + i); + for (i = 1024; i < (5 * 1024); i += 2) - iowrite16(i * 12 + 7, p + i); + iowrite16(v16(i), p + i); + for (i = (5 * 1024); i < (16 * 1024); i += 4) - iowrite32(i * 212371 + 13, p + i); + iowrite32(v32(i), p + i); } static void do_read_test(void __iomem *p) { unsigned int i; + unsigned errs[3] = { 0 }; mmiotrace_printk("Read test.\n"); + for (i = 0; i < 256; i++) - ioread8(p + i); + if (ioread8(p + i) != i) + ++errs[0]; + for (i = 1024; i < (5 * 1024); i += 2) - ioread16(p + i); + if (ioread16(p + i) != v16(i)) + ++errs[1]; + for (i = (5 * 1024); i < (16 * 1024); i += 4) - ioread32(p + i); + if (ioread32(p + i) != v32(i)) + ++errs[2]; + + mmiotrace_printk("Read errors: 8-bit %d, 16-bit %d, 32-bit %d.\n", + errs[0], errs[1], errs[2]); } static void do_test(void) From 5ff93697fcfe1e2b9e61db82961d8f50d1ad5d57 Mon Sep 17 00:00:00 2001 From: Pekka Paalanen Date: Sun, 1 Mar 2009 16:10:08 +0200 Subject: [PATCH 43/92] x86: add far read test to testmmiotrace Apparently pages far into an ioremapped region might not actually be mapped during ioremap(). Add an optional read test to try to trigger a multiply faulting MMIO access. Also add more messages to the kernel log to help debugging. This patch is based on a patch suggested by Stuart Bennett who discovered bugs in mmiotrace related to normal kernel space faults. Signed-off-by: Pekka Paalanen Cc: Stuart Bennett Cc: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/mm/testmmiotrace.c | 35 ++++++++++++++++++++++++++++------- 1 file changed, 28 insertions(+), 7 deletions(-) diff --git a/arch/x86/mm/testmmiotrace.c b/arch/x86/mm/testmmiotrace.c index 4b29f3b0ee01..427fd1b56df5 100644 --- a/arch/x86/mm/testmmiotrace.c +++ b/arch/x86/mm/testmmiotrace.c @@ -9,7 +9,13 @@ static unsigned long mmio_address; module_param(mmio_address, ulong, 0); -MODULE_PARM_DESC(mmio_address, "Start address of the mapping of 16 kB."); +MODULE_PARM_DESC(mmio_address, " Start address of the mapping of 16 kB " + "(or 8 MB if read_far is non-zero)."); + +static unsigned long read_far = 0x400100; +module_param(read_far, ulong, 0); +MODULE_PARM_DESC(read_far, " Offset of a 32-bit read within 8 MB " + "(default: 0x400100)."); static unsigned v16(unsigned i) { @@ -24,6 +30,7 @@ static unsigned v32(unsigned i) static void do_write_test(void __iomem *p) { unsigned int i; + pr_info(MODULE_NAME ": write test.\n"); mmiotrace_printk("Write test.\n"); for (i = 0; i < 256; i++) @@ -40,6 +47,7 @@ static void do_read_test(void __iomem *p) { unsigned int i; unsigned errs[3] = { 0 }; + pr_info(MODULE_NAME ": read test.\n"); mmiotrace_printk("Read test.\n"); for (i = 0; i < 256; i++) @@ -58,9 +66,17 @@ static void do_read_test(void __iomem *p) errs[0], errs[1], errs[2]); } -static void do_test(void) +static void do_read_far_test(void __iomem *p) { - void __iomem *p = ioremap_nocache(mmio_address, 0x4000); + pr_info(MODULE_NAME ": read far test.\n"); + mmiotrace_printk("Read far test.\n"); + + ioread32(p + read_far); +} + +static void do_test(unsigned long size) +{ + void __iomem *p = ioremap_nocache(mmio_address, size); if (!p) { pr_err(MODULE_NAME ": could not ioremap, aborting.\n"); return; @@ -68,11 +84,15 @@ static void do_test(void) mmiotrace_printk("ioremap returned %p.\n", p); do_write_test(p); do_read_test(p); + if (read_far && read_far < size - 4) + do_read_far_test(p); iounmap(p); } static int __init init(void) { + unsigned long size = (read_far) ? (8 << 20) : (16 << 10); + if (mmio_address == 0) { pr_err(MODULE_NAME ": you have to use the module argument " "mmio_address.\n"); @@ -81,10 +101,11 @@ static int __init init(void) return -ENXIO; } - pr_warning(MODULE_NAME ": WARNING: mapping 16 kB @ 0x%08lx " - "in PCI address space, and writing " - "rubbish in there.\n", mmio_address); - do_test(); + pr_warning(MODULE_NAME ": WARNING: mapping %lu kB @ 0x%08lx in PCI " + "address space, and writing 16 kB of rubbish in there.\n", + size >> 10, mmio_address); + do_test(size); + pr_info(MODULE_NAME ": All done.\n"); return 0; } From e9d54cae8f03e7f963a12f44bd50d68f49b9ea36 Mon Sep 17 00:00:00 2001 From: Stuart Bennett Date: Fri, 30 Jan 2009 17:38:59 +0000 Subject: [PATCH 44/92] x86 mmiotrace: WARN_ONCE if dis/arming a page fails Print a full warning once, if arming or disarming a page fails. Also, if initial arming fails, do not handle the page further. This avoids the possibility of a page failing to arm and then later claiming to have handled any fault on that page. WARN_ONCE added by Pekka Paalanen. Signed-off-by: Stuart Bennett Signed-off-by: Pekka Paalanen Cc: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/mm/kmmio.c | 25 +++++++++++++++++-------- 1 file changed, 17 insertions(+), 8 deletions(-) diff --git a/arch/x86/mm/kmmio.c b/arch/x86/mm/kmmio.c index 93d82038af4b..fb1f11546fcd 100644 --- a/arch/x86/mm/kmmio.c +++ b/arch/x86/mm/kmmio.c @@ -105,7 +105,7 @@ static struct kmmio_fault_page *get_kmmio_fault_page(unsigned long page) return NULL; } -static void set_page_present(unsigned long addr, bool present, +static int set_page_present(unsigned long addr, bool present, unsigned int *pglevel) { pteval_t pteval; @@ -116,7 +116,7 @@ static void set_page_present(unsigned long addr, bool present, if (!pte) { pr_err("kmmio: no pte for page 0x%08lx\n", addr); - return; + return -1; } if (pglevel) @@ -140,22 +140,27 @@ static void set_page_present(unsigned long addr, bool present, default: pr_err("kmmio: unexpected page level 0x%x.\n", level); - return; + return -1; } __flush_tlb_one(addr); + + return 0; } /** Mark the given page as not present. Access to it will trigger a fault. */ -static void arm_kmmio_fault_page(unsigned long page, unsigned int *pglevel) +static int arm_kmmio_fault_page(unsigned long page, unsigned int *pglevel) { - set_page_present(page & PAGE_MASK, false, pglevel); + int ret = set_page_present(page & PAGE_MASK, false, pglevel); + WARN_ONCE(ret < 0, KERN_ERR "kmmio arming 0x%08lx failed.\n", page); + return ret; } /** Mark the given page as present. */ static void disarm_kmmio_fault_page(unsigned long page, unsigned int *pglevel) { - set_page_present(page & PAGE_MASK, true, pglevel); + int ret = set_page_present(page & PAGE_MASK, true, pglevel); + WARN_ONCE(ret < 0, KERN_ERR "kmmio disarming 0x%08lx failed.\n", page); } /* @@ -326,9 +331,13 @@ static int add_kmmio_fault_page(unsigned long page) f->count = 1; f->page = page; - list_add_rcu(&f->list, kmmio_page_list(f->page)); - arm_kmmio_fault_page(f->page, NULL); + if (arm_kmmio_fault_page(f->page, NULL)) { + kfree(f); + return -1; + } + + list_add_rcu(&f->list, kmmio_page_list(f->page)); return 0; } From 5359b585fb5edb3db34d6cd491e1475b098c61d3 Mon Sep 17 00:00:00 2001 From: Pekka Paalanen Date: Sun, 1 Mar 2009 16:11:58 +0200 Subject: [PATCH 45/92] x86 mmiotrace: fix save/restore page table state From baa99e2b32449ec7bf147c234adfa444caecac8a Mon Sep 17 00:00:00 2001 From: Pekka Paalanen Date: Sun, 22 Feb 2009 20:02:43 +0200 Blindly setting _PAGE_PRESENT in disarm_kmmio_fault_page() overlooks the possibility, that the page was not present when it was armed. Make arm_kmmio_fault_page() store the previous page presence in struct kmmio_fault_page and use it on disarm. This patch was originally written by Stuart Bennett, but Pekka Paalanen rewrote it a little different. Signed-off-by: Pekka Paalanen Cc: Stuart Bennett Cc: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/mm/kmmio.c | 66 ++++++++++++++++++++++++++++++--------------- 1 file changed, 44 insertions(+), 22 deletions(-) diff --git a/arch/x86/mm/kmmio.c b/arch/x86/mm/kmmio.c index fb1f11546fcd..be361eb828c8 100644 --- a/arch/x86/mm/kmmio.c +++ b/arch/x86/mm/kmmio.c @@ -32,6 +32,8 @@ struct kmmio_fault_page { struct list_head list; struct kmmio_fault_page *release_next; unsigned long page; /* location of the fault page */ + bool old_presence; /* page presence prior to arming */ + bool armed; /* * Number of times this page has been registered as a part @@ -105,8 +107,7 @@ static struct kmmio_fault_page *get_kmmio_fault_page(unsigned long page) return NULL; } -static int set_page_present(unsigned long addr, bool present, - unsigned int *pglevel) +static int set_page_presence(unsigned long addr, bool present, bool *old) { pteval_t pteval; pmdval_t pmdval; @@ -119,20 +120,21 @@ static int set_page_present(unsigned long addr, bool present, return -1; } - if (pglevel) - *pglevel = level; - switch (level) { case PG_LEVEL_2M: pmd = (pmd_t *)pte; - pmdval = pmd_val(*pmd) & ~_PAGE_PRESENT; + pmdval = pmd_val(*pmd); + *old = !!(pmdval & _PAGE_PRESENT); + pmdval &= ~_PAGE_PRESENT; if (present) pmdval |= _PAGE_PRESENT; set_pmd(pmd, __pmd(pmdval)); break; case PG_LEVEL_4K: - pteval = pte_val(*pte) & ~_PAGE_PRESENT; + pteval = pte_val(*pte); + *old = !!(pteval & _PAGE_PRESENT); + pteval &= ~_PAGE_PRESENT; if (present) pteval |= _PAGE_PRESENT; set_pte_atomic(pte, __pte(pteval)); @@ -148,19 +150,39 @@ static int set_page_present(unsigned long addr, bool present, return 0; } -/** Mark the given page as not present. Access to it will trigger a fault. */ -static int arm_kmmio_fault_page(unsigned long page, unsigned int *pglevel) +/* + * Mark the given page as not present. Access to it will trigger a fault. + * + * Struct kmmio_fault_page is protected by RCU and kmmio_lock, but the + * protection is ignored here. RCU read lock is assumed held, so the struct + * will not disappear unexpectedly. Furthermore, the caller must guarantee, + * that double arming the same virtual address (page) cannot occur. + * + * Double disarming on the other hand is allowed, and may occur when a fault + * and mmiotrace shutdown happen simultaneously. + */ +static int arm_kmmio_fault_page(struct kmmio_fault_page *f) { - int ret = set_page_present(page & PAGE_MASK, false, pglevel); - WARN_ONCE(ret < 0, KERN_ERR "kmmio arming 0x%08lx failed.\n", page); + int ret; + WARN_ONCE(f->armed, KERN_ERR "kmmio page already armed.\n"); + if (f->armed) { + pr_warning("kmmio double-arm: page 0x%08lx, ref %d, old %d\n", + f->page, f->count, f->old_presence); + } + ret = set_page_presence(f->page, false, &f->old_presence); + WARN_ONCE(ret < 0, KERN_ERR "kmmio arming 0x%08lx failed.\n", f->page); + f->armed = true; return ret; } -/** Mark the given page as present. */ -static void disarm_kmmio_fault_page(unsigned long page, unsigned int *pglevel) +/** Restore the given page to saved presence state. */ +static void disarm_kmmio_fault_page(struct kmmio_fault_page *f) { - int ret = set_page_present(page & PAGE_MASK, true, pglevel); - WARN_ONCE(ret < 0, KERN_ERR "kmmio disarming 0x%08lx failed.\n", page); + bool tmp; + int ret = set_page_presence(f->page, f->old_presence, &tmp); + WARN_ONCE(ret < 0, + KERN_ERR "kmmio disarming 0x%08lx failed.\n", f->page); + f->armed = false; } /* @@ -207,7 +229,7 @@ int kmmio_handler(struct pt_regs *regs, unsigned long addr) ctx = &get_cpu_var(kmmio_ctx); if (ctx->active) { - disarm_kmmio_fault_page(faultpage->page, NULL); + disarm_kmmio_fault_page(faultpage); if (addr == ctx->addr) { /* * On SMP we sometimes get recursive probe hits on the @@ -249,7 +271,7 @@ int kmmio_handler(struct pt_regs *regs, unsigned long addr) regs->flags &= ~X86_EFLAGS_IF; /* Now we set present bit in PTE and single step. */ - disarm_kmmio_fault_page(ctx->fpage->page, NULL); + disarm_kmmio_fault_page(ctx->fpage); /* * If another cpu accesses the same page while we are stepping, @@ -288,7 +310,7 @@ static int post_kmmio_handler(unsigned long condition, struct pt_regs *regs) if (ctx->probe && ctx->probe->post_handler) ctx->probe->post_handler(ctx->probe, condition, regs); - arm_kmmio_fault_page(ctx->fpage->page, NULL); + arm_kmmio_fault_page(ctx->fpage); regs->flags &= ~X86_EFLAGS_TF; regs->flags |= ctx->saved_flags; @@ -320,19 +342,19 @@ static int add_kmmio_fault_page(unsigned long page) f = get_kmmio_fault_page(page); if (f) { if (!f->count) - arm_kmmio_fault_page(f->page, NULL); + arm_kmmio_fault_page(f); f->count++; return 0; } - f = kmalloc(sizeof(*f), GFP_ATOMIC); + f = kzalloc(sizeof(*f), GFP_ATOMIC); if (!f) return -1; f->count = 1; f->page = page; - if (arm_kmmio_fault_page(f->page, NULL)) { + if (arm_kmmio_fault_page(f)) { kfree(f); return -1; } @@ -356,7 +378,7 @@ static void release_kmmio_fault_page(unsigned long page, f->count--; BUG_ON(f->count < 0); if (!f->count) { - disarm_kmmio_fault_page(f->page, NULL); + disarm_kmmio_fault_page(f); f->release_next = *release_list; *release_list = f; } From 0b700a6a253b6a3b3059bb9a9247a73490ee33fb Mon Sep 17 00:00:00 2001 From: Pekka Paalanen Date: Sun, 1 Mar 2009 16:12:48 +0200 Subject: [PATCH 46/92] x86 mmiotrace: split set_page_presence() From 36772dcb6ffbbb68254cbfc379a103acd2fbfefc Mon Sep 17 00:00:00 2001 From: Pekka Paalanen Date: Sat, 28 Feb 2009 21:34:59 +0200 Split set_page_presence() in kmmio.c into two more functions set_pmd_presence() and set_pte_presence(). Purely code reorganization, no functional changes. Signed-off-by: Pekka Paalanen Cc: Stuart Bennett Cc: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/mm/kmmio.c | 41 ++++++++++++++++++++++------------------- 1 file changed, 22 insertions(+), 19 deletions(-) diff --git a/arch/x86/mm/kmmio.c b/arch/x86/mm/kmmio.c index be361eb828c8..d29777520af3 100644 --- a/arch/x86/mm/kmmio.c +++ b/arch/x86/mm/kmmio.c @@ -107,12 +107,29 @@ static struct kmmio_fault_page *get_kmmio_fault_page(unsigned long page) return NULL; } +static void set_pmd_presence(pmd_t *pmd, bool present, bool *old) +{ + pmdval_t v = pmd_val(*pmd); + *old = !!(v & _PAGE_PRESENT); + v &= ~_PAGE_PRESENT; + if (present) + v |= _PAGE_PRESENT; + set_pmd(pmd, __pmd(v)); +} + +static void set_pte_presence(pte_t *pte, bool present, bool *old) +{ + pteval_t v = pte_val(*pte); + *old = !!(v & _PAGE_PRESENT); + v &= ~_PAGE_PRESENT; + if (present) + v |= _PAGE_PRESENT; + set_pte_atomic(pte, __pte(v)); +} + static int set_page_presence(unsigned long addr, bool present, bool *old) { - pteval_t pteval; - pmdval_t pmdval; unsigned int level; - pmd_t *pmd; pte_t *pte = lookup_address(addr, &level); if (!pte) { @@ -122,31 +139,17 @@ static int set_page_presence(unsigned long addr, bool present, bool *old) switch (level) { case PG_LEVEL_2M: - pmd = (pmd_t *)pte; - pmdval = pmd_val(*pmd); - *old = !!(pmdval & _PAGE_PRESENT); - pmdval &= ~_PAGE_PRESENT; - if (present) - pmdval |= _PAGE_PRESENT; - set_pmd(pmd, __pmd(pmdval)); + set_pmd_presence((pmd_t *)pte, present, old); break; - case PG_LEVEL_4K: - pteval = pte_val(*pte); - *old = !!(pteval & _PAGE_PRESENT); - pteval &= ~_PAGE_PRESENT; - if (present) - pteval |= _PAGE_PRESENT; - set_pte_atomic(pte, __pte(pteval)); + set_pte_presence(pte, present, old); break; - default: pr_err("kmmio: unexpected page level 0x%x.\n", level); return -1; } __flush_tlb_one(addr); - return 0; } From 3e39aa156a24ce386da378784edd0f748c770087 Mon Sep 17 00:00:00 2001 From: Stuart Bennett Date: Thu, 5 Feb 2009 11:02:02 +0000 Subject: [PATCH 47/92] x86 mmiotrace: improve handling of secondary faults Upgrade some kmmio.c debug messages to warnings. Allow secondary faults on probed pages to fall through, and only log secondary faults that are not due to non-present pages. Patch edited by Pekka Paalanen. Signed-off-by: Stuart Bennett Signed-off-by: Pekka Paalanen Cc: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/mm/kmmio.c | 40 ++++++++++++++++++++++------------------ 1 file changed, 22 insertions(+), 18 deletions(-) diff --git a/arch/x86/mm/kmmio.c b/arch/x86/mm/kmmio.c index d29777520af3..4c66bd3a240d 100644 --- a/arch/x86/mm/kmmio.c +++ b/arch/x86/mm/kmmio.c @@ -232,28 +232,32 @@ int kmmio_handler(struct pt_regs *regs, unsigned long addr) ctx = &get_cpu_var(kmmio_ctx); if (ctx->active) { - disarm_kmmio_fault_page(faultpage); if (addr == ctx->addr) { /* - * On SMP we sometimes get recursive probe hits on the - * same address. Context is already saved, fall out. + * A second fault on the same page means some other + * condition needs handling by do_page_fault(), the + * page really not being present is the most common. */ - pr_debug("kmmio: duplicate probe hit on CPU %d, for " - "address 0x%08lx.\n", - smp_processor_id(), addr); - ret = 1; - goto no_kmmio_ctx; - } - /* - * Prevent overwriting already in-flight context. - * This should not happen, let's hope disarming at least - * prevents a panic. - */ - pr_emerg("kmmio: recursive probe hit on CPU %d, " + pr_debug("kmmio: secondary hit for 0x%08lx CPU %d.\n", + addr, smp_processor_id()); + + if (!faultpage->old_presence) + pr_info("kmmio: unexpected secondary hit for " + "address 0x%08lx on CPU %d.\n", addr, + smp_processor_id()); + } else { + /* + * Prevent overwriting already in-flight context. + * This should not happen, let's hope disarming at + * least prevents a panic. + */ + pr_emerg("kmmio: recursive probe hit on CPU %d, " "for address 0x%08lx. Ignoring.\n", smp_processor_id(), addr); - pr_emerg("kmmio: previous hit was at 0x%08lx.\n", - ctx->addr); + pr_emerg("kmmio: previous hit was at 0x%08lx.\n", + ctx->addr); + disarm_kmmio_fault_page(faultpage); + } goto no_kmmio_ctx; } ctx->active++; @@ -305,7 +309,7 @@ static int post_kmmio_handler(unsigned long condition, struct pt_regs *regs) struct kmmio_context *ctx = &get_cpu_var(kmmio_ctx); if (!ctx->active) { - pr_debug("kmmio: spurious debug trap on CPU %d.\n", + pr_warning("kmmio: spurious debug trap on CPU %d.\n", smp_processor_id()); goto out; } From 340430c572f7b2b275d39965e88bafa71693cb23 Mon Sep 17 00:00:00 2001 From: Pekka Paalanen Date: Tue, 24 Feb 2009 21:44:15 +0200 Subject: [PATCH 48/92] x86 mmiotrace: fix race with release_kmmio_fault_page() There was a theoretical possibility to a race between arming a page in post_kmmio_handler() and disarming the page in release_kmmio_fault_page(): cpu0 cpu1 ------------------------------------------------------------------ mmiotrace shutdown enter release_kmmio_fault_page fault on the page disarm the page disarm the page handle the MMIO access re-arm the page put the page on release list remove_kmmio_fault_pages() fault on the page page not known to mmiotrace fall back to do_page_fault() *KABOOM* (This scenario also shows the double disarm case which is allowed.) Fixed by acquiring kmmio_lock in post_kmmio_handler() and checking if the page is being released from mmiotrace. Signed-off-by: Pekka Paalanen Cc: Stuart Bennett Cc: Steven Rostedt Signed-off-by: Ingo Molnar --- arch/x86/mm/kmmio.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/arch/x86/mm/kmmio.c b/arch/x86/mm/kmmio.c index 4c66bd3a240d..9f205030d9aa 100644 --- a/arch/x86/mm/kmmio.c +++ b/arch/x86/mm/kmmio.c @@ -38,7 +38,8 @@ struct kmmio_fault_page { /* * Number of times this page has been registered as a part * of a probe. If zero, page is disarmed and this may be freed. - * Used only by writers (RCU). + * Used only by writers (RCU) and post_kmmio_handler(). + * Protected by kmmio_lock, when linked into kmmio_page_table. */ int count; }; @@ -317,7 +318,11 @@ static int post_kmmio_handler(unsigned long condition, struct pt_regs *regs) if (ctx->probe && ctx->probe->post_handler) ctx->probe->post_handler(ctx->probe, condition, regs); - arm_kmmio_fault_page(ctx->fpage); + /* Prevent racing against release_kmmio_fault_page(). */ + spin_lock(&kmmio_lock); + if (ctx->fpage->count) + arm_kmmio_fault_page(ctx->fpage); + spin_unlock(&kmmio_lock); regs->flags &= ~X86_EFLAGS_TF; regs->flags |= ctx->saved_flags; From f945405cdecd9e0ae3e58ff84cabd19b4522965e Mon Sep 17 00:00:00 2001 From: Ben Dooks Date: Fri, 20 Feb 2009 20:33:08 +0300 Subject: [PATCH 49/92] sdhci: Add quirk for controllers with no end-of-busy IRQ The Samsung SDHCI (and FSL eSDHC) controller block seems to fail to generate an INT_DATA_END after the transfer has completed and the bus busy state finished. Changes in e809517f6fa5803a5a1cd56026f0e2190fc13d5c to use the new busy method are the cause of the behaviour change. Signed-off-by: Ben Dooks Signed-off-by: Anton Vorontsov Signed-off-by: Pierre Ossman --- drivers/mmc/host/sdhci.c | 5 ++++- drivers/mmc/host/sdhci.h | 2 ++ 2 files changed, 6 insertions(+), 1 deletion(-) diff --git a/drivers/mmc/host/sdhci.c b/drivers/mmc/host/sdhci.c index f52f3053ed92..accb592764ed 100644 --- a/drivers/mmc/host/sdhci.c +++ b/drivers/mmc/host/sdhci.c @@ -1291,8 +1291,11 @@ static void sdhci_cmd_irq(struct sdhci_host *host, u32 intmask) if (host->cmd->data) DBG("Cannot wait for busy signal when also " "doing a data transfer"); - else + else if (!(host->quirks & SDHCI_QUIRK_NO_BUSY_IRQ)) return; + + /* The controller does not support the end-of-busy IRQ, + * fall through and take the SDHCI_INT_RESPONSE */ } if (intmask & SDHCI_INT_RESPONSE) diff --git a/drivers/mmc/host/sdhci.h b/drivers/mmc/host/sdhci.h index ebb83657e27a..43c37c68d07a 100644 --- a/drivers/mmc/host/sdhci.h +++ b/drivers/mmc/host/sdhci.h @@ -208,6 +208,8 @@ struct sdhci_host { #define SDHCI_QUIRK_BROKEN_TIMEOUT_VAL (1<<12) /* Controller has an issue with buffer bits for small transfers */ #define SDHCI_QUIRK_BROKEN_SMALL_PIO (1<<13) +/* Controller does not provide transfer-complete interrupt when not busy */ +#define SDHCI_QUIRK_NO_BUSY_IRQ (1<<14) int irq; /* Device IRQ */ void __iomem * ioaddr; /* Mapped address */ From a0874897b1ba106298e4303a25456a473fc40f3d Mon Sep 17 00:00:00 2001 From: Andres Salomon Date: Mon, 2 Mar 2009 21:48:20 +0100 Subject: [PATCH 50/92] sdhci: Add NO_BUSY_IRQ quirk for Marvell CAFE host chip As described here: http://lkml.org/lkml/2009/2/20/265 The CAFE chip is broken due to commit e809517f6fa5803a5a1cd5602. Anton added a quirk here: http://lkml.org/lkml/2009/2/20/279 that fixes CAFE's problem. This adds the quirk for CAFE. Signed-off-by: Andres Salomon Signed-off-by: Pierre Ossman --- drivers/mmc/host/sdhci-pci.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/mmc/host/sdhci-pci.c b/drivers/mmc/host/sdhci-pci.c index 8cff5f5e7f86..406da9a8d453 100644 --- a/drivers/mmc/host/sdhci-pci.c +++ b/drivers/mmc/host/sdhci-pci.c @@ -107,6 +107,7 @@ static const struct sdhci_pci_fixes sdhci_ene_714 = { static const struct sdhci_pci_fixes sdhci_cafe = { .quirks = SDHCI_QUIRK_NO_SIMULT_VDD_AND_POWER | + SDHCI_QUIRK_NO_BUSY_IRQ | SDHCI_QUIRK_BROKEN_TIMEOUT_VAL, }; From ccbe495caa5e604b04d5a31d7459a6f6a76a756c Mon Sep 17 00:00:00 2001 From: Roland McGrath Date: Fri, 27 Feb 2009 19:03:24 -0800 Subject: [PATCH 51/92] x86-64: syscall-audit: fix 32/64 syscall hole On x86-64, a 32-bit process (TIF_IA32) can switch to 64-bit mode with ljmp, and then use the "syscall" instruction to make a 64-bit system call. A 64-bit process make a 32-bit system call with int $0x80. In both these cases, audit_syscall_entry() will use the wrong system call number table and the wrong system call argument registers. This could be used to circumvent a syscall audit configuration that filters based on the syscall numbers or argument details. Signed-off-by: Roland McGrath Signed-off-by: Linus Torvalds --- arch/x86/kernel/ptrace.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/arch/x86/kernel/ptrace.c b/arch/x86/kernel/ptrace.c index 5a4c23d89892..06ca07f6ad86 100644 --- a/arch/x86/kernel/ptrace.c +++ b/arch/x86/kernel/ptrace.c @@ -1388,7 +1388,7 @@ void send_sigtrap(struct task_struct *tsk, struct pt_regs *regs, #ifdef CONFIG_X86_32 # define IS_IA32 1 #elif defined CONFIG_IA32_EMULATION -# define IS_IA32 test_thread_flag(TIF_IA32) +# define IS_IA32 is_compat_task() #else # define IS_IA32 0 #endif From 5b1017404aea6d2e552e991b3fd814d839e9cd67 Mon Sep 17 00:00:00 2001 From: Roland McGrath Date: Fri, 27 Feb 2009 23:25:54 -0800 Subject: [PATCH 52/92] x86-64: seccomp: fix 32/64 syscall hole On x86-64, a 32-bit process (TIF_IA32) can switch to 64-bit mode with ljmp, and then use the "syscall" instruction to make a 64-bit system call. A 64-bit process make a 32-bit system call with int $0x80. In both these cases under CONFIG_SECCOMP=y, secure_computing() will use the wrong system call number table. The fix is simple: test TS_COMPAT instead of TIF_IA32. Here is an example exploit: /* test case for seccomp circumvention on x86-64 There are two failure modes: compile with -m64 or compile with -m32. The -m64 case is the worst one, because it does "chmod 777 ." (could be any chmod call). The -m32 case demonstrates it was able to do stat(), which can glean information but not harm anything directly. A buggy kernel will let the test do something, print, and exit 1; a fixed kernel will make it exit with SIGKILL before it does anything. */ #define _GNU_SOURCE #include #include #include #include #include #include #include int main (int argc, char **argv) { char buf[100]; static const char dot[] = "."; long ret; unsigned st[24]; if (prctl (PR_SET_SECCOMP, 1, 0, 0, 0) != 0) perror ("prctl(PR_SET_SECCOMP) -- not compiled into kernel?"); #ifdef __x86_64__ assert ((uintptr_t) dot < (1UL << 32)); asm ("int $0x80 # %0 <- %1(%2 %3)" : "=a" (ret) : "0" (15), "b" (dot), "c" (0777)); ret = snprintf (buf, sizeof buf, "result %ld (check mode on .!)\n", ret); #elif defined __i386__ asm (".code32\n" "pushl %%cs\n" "pushl $2f\n" "ljmpl $0x33, $1f\n" ".code64\n" "1: syscall # %0 <- %1(%2 %3)\n" "lretl\n" ".code32\n" "2:" : "=a" (ret) : "0" (4), "D" (dot), "S" (&st)); if (ret == 0) ret = snprintf (buf, sizeof buf, "stat . -> st_uid=%u\n", st[7]); else ret = snprintf (buf, sizeof buf, "result %ld\n", ret); #else # error "not this one" #endif write (1, buf, ret); syscall (__NR_exit, 1); return 2; } Signed-off-by: Roland McGrath [ I don't know if anybody actually uses seccomp, but it's enabled in at least both Fedora and SuSE kernels, so maybe somebody is. - Linus ] Signed-off-by: Linus Torvalds --- arch/mips/include/asm/seccomp.h | 1 - arch/powerpc/include/asm/compat.h | 5 +++++ arch/powerpc/include/asm/seccomp.h | 4 ---- arch/sparc/include/asm/compat.h | 5 +++++ arch/sparc/include/asm/seccomp.h | 6 ------ arch/x86/include/asm/seccomp_32.h | 6 ------ arch/x86/include/asm/seccomp_64.h | 8 -------- kernel/seccomp.c | 7 ++++--- 8 files changed, 14 insertions(+), 28 deletions(-) diff --git a/arch/mips/include/asm/seccomp.h b/arch/mips/include/asm/seccomp.h index 36ed44070256..a6772e9507f5 100644 --- a/arch/mips/include/asm/seccomp.h +++ b/arch/mips/include/asm/seccomp.h @@ -1,6 +1,5 @@ #ifndef __ASM_SECCOMP_H -#include #include #define __NR_seccomp_read __NR_read diff --git a/arch/powerpc/include/asm/compat.h b/arch/powerpc/include/asm/compat.h index d811a8cd7b58..4774c2f92232 100644 --- a/arch/powerpc/include/asm/compat.h +++ b/arch/powerpc/include/asm/compat.h @@ -210,5 +210,10 @@ struct compat_shmid64_ds { compat_ulong_t __unused6; }; +static inline int is_compat_task(void) +{ + return test_thread_flag(TIF_32BIT); +} + #endif /* __KERNEL__ */ #endif /* _ASM_POWERPC_COMPAT_H */ diff --git a/arch/powerpc/include/asm/seccomp.h b/arch/powerpc/include/asm/seccomp.h index 853765eb1f65..00c1d9133cfe 100644 --- a/arch/powerpc/include/asm/seccomp.h +++ b/arch/powerpc/include/asm/seccomp.h @@ -1,10 +1,6 @@ #ifndef _ASM_POWERPC_SECCOMP_H #define _ASM_POWERPC_SECCOMP_H -#ifdef __KERNEL__ -#include -#endif - #include #define __NR_seccomp_read __NR_read diff --git a/arch/sparc/include/asm/compat.h b/arch/sparc/include/asm/compat.h index f260b58f5ce9..0e706257918f 100644 --- a/arch/sparc/include/asm/compat.h +++ b/arch/sparc/include/asm/compat.h @@ -240,4 +240,9 @@ struct compat_shmid64_ds { unsigned int __unused2; }; +static inline int is_compat_task(void) +{ + return test_thread_flag(TIF_32BIT); +} + #endif /* _ASM_SPARC64_COMPAT_H */ diff --git a/arch/sparc/include/asm/seccomp.h b/arch/sparc/include/asm/seccomp.h index 7fcd9968192b..adca1bce41d4 100644 --- a/arch/sparc/include/asm/seccomp.h +++ b/arch/sparc/include/asm/seccomp.h @@ -1,11 +1,5 @@ #ifndef _ASM_SECCOMP_H -#include /* already defines TIF_32BIT */ - -#ifndef TIF_32BIT -#error "unexpected TIF_32BIT on sparc64" -#endif - #include #define __NR_seccomp_read __NR_read diff --git a/arch/x86/include/asm/seccomp_32.h b/arch/x86/include/asm/seccomp_32.h index a6ad87b352c4..b811d6f5780c 100644 --- a/arch/x86/include/asm/seccomp_32.h +++ b/arch/x86/include/asm/seccomp_32.h @@ -1,12 +1,6 @@ #ifndef _ASM_X86_SECCOMP_32_H #define _ASM_X86_SECCOMP_32_H -#include - -#ifdef TIF_32BIT -#error "unexpected TIF_32BIT on i386" -#endif - #include #define __NR_seccomp_read __NR_read diff --git a/arch/x86/include/asm/seccomp_64.h b/arch/x86/include/asm/seccomp_64.h index 4171bb794e9e..84ec1bd161a5 100644 --- a/arch/x86/include/asm/seccomp_64.h +++ b/arch/x86/include/asm/seccomp_64.h @@ -1,14 +1,6 @@ #ifndef _ASM_X86_SECCOMP_64_H #define _ASM_X86_SECCOMP_64_H -#include - -#ifdef TIF_32BIT -#error "unexpected TIF_32BIT on x86_64" -#else -#define TIF_32BIT TIF_IA32 -#endif - #include #include diff --git a/kernel/seccomp.c b/kernel/seccomp.c index ad64fcb731f2..57d4b13b631d 100644 --- a/kernel/seccomp.c +++ b/kernel/seccomp.c @@ -8,6 +8,7 @@ #include #include +#include /* #define SECCOMP_DEBUG 1 */ #define NR_SECCOMP_MODES 1 @@ -22,7 +23,7 @@ static int mode1_syscalls[] = { 0, /* null terminated */ }; -#ifdef TIF_32BIT +#ifdef CONFIG_COMPAT static int mode1_syscalls_32[] = { __NR_seccomp_read_32, __NR_seccomp_write_32, __NR_seccomp_exit_32, __NR_seccomp_sigreturn_32, 0, /* null terminated */ @@ -37,8 +38,8 @@ void __secure_computing(int this_syscall) switch (mode) { case 1: syscall = mode1_syscalls; -#ifdef TIF_32BIT - if (test_thread_flag(TIF_32BIT)) +#ifdef CONFIG_COMPAT + if (is_compat_task()) syscall = mode1_syscalls_32; #endif do { From fbfd8b5622a18afefb53a69d31fbddecba404f64 Mon Sep 17 00:00:00 2001 From: Karsten Keil Date: Sun, 1 Mar 2009 18:04:53 +0100 Subject: [PATCH 53/92] Change email address Since I will loose the old address soon, please change it. Signed-off-by: Karsten Keil Signed-off-by: Linus Torvalds --- MAINTAINERS | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/MAINTAINERS b/MAINTAINERS index 59fd2d1d94a7..1c2ca1dc66f2 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -2464,7 +2464,7 @@ S: Maintained ISDN SUBSYSTEM P: Karsten Keil -M: kkeil@suse.de +M: isdn@linux-pingi.de L: isdn4linux@listserv.isdn4linux.de (subscribers-only) W: http://www.isdn4linux.de T: git kernel.org:/pub/scm/linux/kernel/kkeil/isdn-2.6.git From 155b25bcc28631a5b5230191aa3f56c40dfffa3f Mon Sep 17 00:00:00 2001 From: Randy Dunlap Date: Mon, 2 Mar 2009 14:14:06 -0800 Subject: [PATCH 54/92] menu: fix embedded menu snafu The COMPAT_BRK kconfig symbol does not depend on EMBEDDED, but it is in the midst of the EMBEDDED menu symbols, so it mucks up the EMBEDDED menu. Fix by moving it to just after all of the EMBEDDED menu symbols. Also, surround all of the EMBEDDED symbols with "if EMBEDDED"/"endif" so that this EMBEDDED block is clearer. Signed-off-by: Randy Dunlap Signed-off-by: Linus Torvalds --- init/Kconfig | 28 ++++++++++++++++------------ 1 file changed, 16 insertions(+), 12 deletions(-) diff --git a/init/Kconfig b/init/Kconfig index f068071fcc5d..90003163e557 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -683,6 +683,8 @@ menuconfig EMBEDDED environments which can tolerate a "non-standard" kernel. Only use this if you really know what you are doing. +if EMBEDDED + config UID16 bool "Enable 16-bit UID system calls" if EMBEDDED depends on ARM || BLACKFIN || CRIS || FRV || H8300 || X86_32 || M68K || (S390 && !64BIT) || SUPERH || SPARC32 || (SPARC64 && COMPAT) || UML || (X86_64 && IA32_EMULATION) @@ -780,18 +782,6 @@ config PCSPKR_PLATFORM This option allows to disable the internal PC-Speaker support, saving some memory. -config COMPAT_BRK - bool "Disable heap randomization" - default y - help - Randomizing heap placement makes heap exploits harder, but it - also breaks ancient binaries (including anything libc5 based). - This option changes the bootup default to heap randomization - disabled, and can be overriden runtime by setting - /proc/sys/kernel/randomize_va_space to 2. - - On non-ancient distros (post-2000 ones) N is usually a safe choice. - config BASE_FULL default y bool "Enable full-sized data structures for core" if EMBEDDED @@ -897,6 +887,20 @@ config SLUB_DEBUG SLUB sysfs support. /sys/slab will not exist and there will be no support for cache validation etc. +endif # EMBEDDED + +config COMPAT_BRK + bool "Disable heap randomization" + default y + help + Randomizing heap placement makes heap exploits harder, but it + also breaks ancient binaries (including anything libc5 based). + This option changes the bootup default to heap randomization + disabled, and can be overriden runtime by setting + /proc/sys/kernel/randomize_va_space to 2. + + On non-ancient distros (post-2000 ones) N is usually a safe choice. + choice prompt "Choose SLAB allocator" default SLUB From 4d77c88e912e5eb9480432af09e950ca8995c253 Mon Sep 17 00:00:00 2001 From: Thomas Hellstrom Date: Mon, 2 Mar 2009 11:10:54 +0100 Subject: [PATCH 55/92] drm: Don't return ERESTARTSYS to user-space. That return code is for in-kernel use only. Use EINTR instead. Signed-off-by: Thomas Hellstrom Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_lock.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_lock.c b/drivers/gpu/drm/drm_lock.c index 46e7b28f0707..b03586fdabbd 100644 --- a/drivers/gpu/drm/drm_lock.c +++ b/drivers/gpu/drm/drm_lock.c @@ -93,7 +93,7 @@ int drm_lock(struct drm_device *dev, void *data, struct drm_file *file_priv) /* Contention */ schedule(); if (signal_pending(current)) { - ret = -ERESTARTSYS; + ret = -EINTR; break; } } From 171901d15deeef61aa8e1b0d0772404f39691b73 Mon Sep 17 00:00:00 2001 From: Thomas Hellstrom Date: Mon, 2 Mar 2009 11:10:55 +0100 Subject: [PATCH 56/92] drm: Wake up all lock waiters when the master disappears. Currently only one waiter is woken up, leaving other waiters hanging waiting for the DRM lock. Signed-off-by: Thomas Hellstrom Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_bufs.c | 2 +- drivers/gpu/drm/drm_stub.c | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/drm_bufs.c b/drivers/gpu/drm/drm_bufs.c index 72c667f9bee1..12715d3c078d 100644 --- a/drivers/gpu/drm/drm_bufs.c +++ b/drivers/gpu/drm/drm_bufs.c @@ -420,7 +420,7 @@ int drm_rmmap_locked(struct drm_device *dev, drm_local_map_t *map) dev->sigdata.lock = NULL; master->lock.hw_lock = NULL; /* SHM removed */ master->lock.file_priv = NULL; - wake_up_interruptible(&master->lock.lock_queue); + wake_up_interruptible_all(&master->lock.lock_queue); } break; case _DRM_AGP: diff --git a/drivers/gpu/drm/drm_stub.c b/drivers/gpu/drm/drm_stub.c index 46bb923b097c..7b251dd70811 100644 --- a/drivers/gpu/drm/drm_stub.c +++ b/drivers/gpu/drm/drm_stub.c @@ -151,7 +151,7 @@ static void drm_master_destroy(struct kref *kref) dev->sigdata.lock = NULL; master->lock.hw_lock = NULL; master->lock.file_priv = NULL; - wake_up_interruptible(&master->lock.lock_queue); + wake_up_interruptible_all(&master->lock.lock_queue); } drm_free(master, sizeof(*master), DRM_MEM_DRIVER); From fda714c29cdf360464059044b221450decb4b913 Mon Sep 17 00:00:00 2001 From: Thomas Hellstrom Date: Mon, 2 Mar 2009 11:10:56 +0100 Subject: [PATCH 57/92] drm: Avoid client deadlocks when the master disappears. This is done by 1) Wake up lock waiters when we close the master file descriptor. Not when the master structure is removed, since the latter requires the waiters themselves to release the refcount on the master structure -> Deadlock. 2) Send a SIGTERM to all clients waiting for the lock. Normally these clients will get a SIGPIPE when the X server dies, but clients may also spin trying to grab the DRM lock, without getting any sort of notification. Signed-off-by: Thomas Hellstrom Signed-off-by: Dave Airlie --- drivers/gpu/drm/drm_fops.c | 14 ++++++++++++++ drivers/gpu/drm/drm_lock.c | 1 + drivers/gpu/drm/drm_stub.c | 8 -------- 3 files changed, 15 insertions(+), 8 deletions(-) diff --git a/drivers/gpu/drm/drm_fops.c b/drivers/gpu/drm/drm_fops.c index 6c020fe5431c..f52663ebe016 100644 --- a/drivers/gpu/drm/drm_fops.c +++ b/drivers/gpu/drm/drm_fops.c @@ -484,6 +484,7 @@ int drm_release(struct inode *inode, struct file *filp) mutex_lock(&dev->struct_mutex); if (file_priv->is_master) { + struct drm_master *master = file_priv->master; struct drm_file *temp; list_for_each_entry(temp, &dev->filelist, lhead) { if ((temp->master == file_priv->master) && @@ -491,6 +492,19 @@ int drm_release(struct inode *inode, struct file *filp) temp->authenticated = 0; } + /** + * Since the master is disappearing, so is the + * possibility to lock. + */ + + if (master->lock.hw_lock) { + if (dev->sigdata.lock == master->lock.hw_lock) + dev->sigdata.lock = NULL; + master->lock.hw_lock = NULL; + master->lock.file_priv = NULL; + wake_up_interruptible_all(&master->lock.lock_queue); + } + if (file_priv->minor->master == file_priv->master) { /* drop the reference held my the minor */ drm_master_put(&file_priv->minor->master); diff --git a/drivers/gpu/drm/drm_lock.c b/drivers/gpu/drm/drm_lock.c index b03586fdabbd..e2f70a516c34 100644 --- a/drivers/gpu/drm/drm_lock.c +++ b/drivers/gpu/drm/drm_lock.c @@ -80,6 +80,7 @@ int drm_lock(struct drm_device *dev, void *data, struct drm_file *file_priv) __set_current_state(TASK_INTERRUPTIBLE); if (!master->lock.hw_lock) { /* Device has been unregistered */ + send_sig(SIGTERM, current, 0); ret = -EINTR; break; } diff --git a/drivers/gpu/drm/drm_stub.c b/drivers/gpu/drm/drm_stub.c index 7b251dd70811..096e2a37446d 100644 --- a/drivers/gpu/drm/drm_stub.c +++ b/drivers/gpu/drm/drm_stub.c @@ -146,14 +146,6 @@ static void drm_master_destroy(struct kref *kref) drm_ht_remove(&master->magiclist); - if (master->lock.hw_lock) { - if (dev->sigdata.lock == master->lock.hw_lock) - dev->sigdata.lock = NULL; - master->lock.hw_lock = NULL; - master->lock.file_priv = NULL; - wake_up_interruptible_all(&master->lock.lock_queue); - } - drm_free(master, sizeof(*master), DRM_MEM_DRIVER); } From 299eb93c5f651b2bc368ada67d8471e4c575fa21 Mon Sep 17 00:00:00 2001 From: Eric Anholt Date: Tue, 24 Feb 2009 22:14:12 -0800 Subject: [PATCH 58/92] drm/i915: Fix use-before-null-check in i915_irq_emit(). This could be triggered by a client asking to emit an irq when the device wasn't initialized. Signed-off-by: Eric Anholt Signed-off-by: Dave Airlie --- drivers/gpu/drm/i915/i915_irq.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/i915/i915_irq.c b/drivers/gpu/drm/i915/i915_irq.c index 548ff2c66431..87b6b603469e 100644 --- a/drivers/gpu/drm/i915/i915_irq.c +++ b/drivers/gpu/drm/i915/i915_irq.c @@ -383,12 +383,13 @@ int i915_irq_emit(struct drm_device *dev, void *data, drm_i915_irq_emit_t *emit = data; int result; - RING_LOCK_TEST_WITH_RETURN(dev, file_priv); - if (!dev_priv) { DRM_ERROR("called with no initialization\n"); return -EINVAL; } + + RING_LOCK_TEST_WITH_RETURN(dev, file_priv); + mutex_lock(&dev->struct_mutex); result = i915_emit_irq(dev); mutex_unlock(&dev->struct_mutex); From 2450cf51a1bdba7037e91b1bcc494b01c58aaf66 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Mon, 2 Mar 2009 16:23:33 -0800 Subject: [PATCH 59/92] Revert "menu: fix embedded menu snafu" This reverts commit 155b25bcc28631a5b5230191aa3f56c40dfffa3f, which was totally wrong - the "embedded" options still exists (very much so) even on non-embedded platforms. It's just that we don't bother with actually asking about them when we're not embedded, we just take their default values (which is usually 'y' - the options add features that may not be worth it in a constrained environment). Noticed-by: Andrew Morton Cc: Randy Dunlap Signed-off-by: Linus Torvalds --- init/Kconfig | 28 ++++++++++++---------------- 1 file changed, 12 insertions(+), 16 deletions(-) diff --git a/init/Kconfig b/init/Kconfig index 90003163e557..f068071fcc5d 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -683,8 +683,6 @@ menuconfig EMBEDDED environments which can tolerate a "non-standard" kernel. Only use this if you really know what you are doing. -if EMBEDDED - config UID16 bool "Enable 16-bit UID system calls" if EMBEDDED depends on ARM || BLACKFIN || CRIS || FRV || H8300 || X86_32 || M68K || (S390 && !64BIT) || SUPERH || SPARC32 || (SPARC64 && COMPAT) || UML || (X86_64 && IA32_EMULATION) @@ -782,6 +780,18 @@ config PCSPKR_PLATFORM This option allows to disable the internal PC-Speaker support, saving some memory. +config COMPAT_BRK + bool "Disable heap randomization" + default y + help + Randomizing heap placement makes heap exploits harder, but it + also breaks ancient binaries (including anything libc5 based). + This option changes the bootup default to heap randomization + disabled, and can be overriden runtime by setting + /proc/sys/kernel/randomize_va_space to 2. + + On non-ancient distros (post-2000 ones) N is usually a safe choice. + config BASE_FULL default y bool "Enable full-sized data structures for core" if EMBEDDED @@ -887,20 +897,6 @@ config SLUB_DEBUG SLUB sysfs support. /sys/slab will not exist and there will be no support for cache validation etc. -endif # EMBEDDED - -config COMPAT_BRK - bool "Disable heap randomization" - default y - help - Randomizing heap placement makes heap exploits harder, but it - also breaks ancient binaries (including anything libc5 based). - This option changes the bootup default to heap randomization - disabled, and can be overriden runtime by setting - /proc/sys/kernel/randomize_va_space to 2. - - On non-ancient distros (post-2000 ones) N is usually a safe choice. - choice prompt "Choose SLAB allocator" default SLUB From d1dd524785e30cf3d64d395d829b207376acb0aa Mon Sep 17 00:00:00 2001 From: Vlad Yasevich Date: Mon, 2 Mar 2009 06:46:50 +0000 Subject: [PATCH 60/92] sctp: fix crash during module unload An extra list_del() during the module load failure and unload resulted in a crash with a list corruption. Now sctp can be unloaded again. Signed-off-by: Vlad Yasevich Signed-off-by: David S. Miller --- net/sctp/protocol.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index b78e3be69013..4e6638449639 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -1322,9 +1322,8 @@ SCTP_STATIC __init int sctp_init(void) out: return status; err_v6_add_protocol: - sctp_v6_del_protocol(); -err_add_protocol: sctp_v4_del_protocol(); +err_add_protocol: inet_ctl_sock_destroy(sctp_ctl_sock); err_ctl_sock_init: sctp_v6_protosw_exit(); @@ -1335,7 +1334,6 @@ err_protosw_init: sctp_v4_pf_exit(); sctp_v6_pf_exit(); sctp_sysctl_unregister(); - list_del(&sctp_af_inet.list); free_pages((unsigned long)sctp_port_hashtable, get_order(sctp_port_hashsize * sizeof(struct sctp_bind_hashbucket))); @@ -1383,7 +1381,6 @@ SCTP_STATIC __exit void sctp_exit(void) sctp_v4_pf_exit(); sctp_sysctl_unregister(); - list_del(&sctp_af_inet.list); free_pages((unsigned long)sctp_assoc_hashtable, get_order(sctp_assoc_hashsize * From 3df2678737974accf437dad11e584c1871a3ede3 Mon Sep 17 00:00:00 2001 From: Wei Yongjun Date: Mon, 2 Mar 2009 06:46:51 +0000 Subject: [PATCH 61/92] sctp: fix kernel panic with ERROR chunk containing too many error causes If ERROR chunk is received with too many error causes in ESTABLISHED state, the kernel get panic. This is because sctp limit the max length of cmds to 14, but while ERROR chunk is received, one error cause will add around 2 cmds by sctp_add_cmd_sf(). So many error causes will fill the limit of cmds and panic. This patch fixed the problem. This bug can be test by SCTP Conformance Test Suite . Signed-off-by: Wei Yongjun Signed-off-by: Vlad Yasevich Signed-off-by: David S. Miller --- net/sctp/sm_sideeffect.c | 54 ++++++++++++++++++++++++---------------- net/sctp/sm_statefuns.c | 16 ++---------- 2 files changed, 35 insertions(+), 35 deletions(-) diff --git a/net/sctp/sm_sideeffect.c b/net/sctp/sm_sideeffect.c index e1d6076b4f59..b5495aecab60 100644 --- a/net/sctp/sm_sideeffect.c +++ b/net/sctp/sm_sideeffect.c @@ -787,36 +787,48 @@ static void sctp_cmd_process_operr(sctp_cmd_seq_t *cmds, struct sctp_association *asoc, struct sctp_chunk *chunk) { - struct sctp_operr_chunk *operr_chunk; struct sctp_errhdr *err_hdr; + struct sctp_ulpevent *ev; - operr_chunk = (struct sctp_operr_chunk *)chunk->chunk_hdr; - err_hdr = &operr_chunk->err_hdr; + while (chunk->chunk_end > chunk->skb->data) { + err_hdr = (struct sctp_errhdr *)(chunk->skb->data); - switch (err_hdr->cause) { - case SCTP_ERROR_UNKNOWN_CHUNK: - { - struct sctp_chunkhdr *unk_chunk_hdr; + ev = sctp_ulpevent_make_remote_error(asoc, chunk, 0, + GFP_ATOMIC); + if (!ev) + return; - unk_chunk_hdr = (struct sctp_chunkhdr *)err_hdr->variable; - switch (unk_chunk_hdr->type) { - /* ADDIP 4.1 A9) If the peer responds to an ASCONF with an - * ERROR chunk reporting that it did not recognized the ASCONF - * chunk type, the sender of the ASCONF MUST NOT send any - * further ASCONF chunks and MUST stop its T-4 timer. - */ - case SCTP_CID_ASCONF: - asoc->peer.asconf_capable = 0; - sctp_add_cmd_sf(cmds, SCTP_CMD_TIMER_STOP, + sctp_ulpq_tail_event(&asoc->ulpq, ev); + + switch (err_hdr->cause) { + case SCTP_ERROR_UNKNOWN_CHUNK: + { + sctp_chunkhdr_t *unk_chunk_hdr; + + unk_chunk_hdr = (sctp_chunkhdr_t *)err_hdr->variable; + switch (unk_chunk_hdr->type) { + /* ADDIP 4.1 A9) If the peer responds to an ASCONF with + * an ERROR chunk reporting that it did not recognized + * the ASCONF chunk type, the sender of the ASCONF MUST + * NOT send any further ASCONF chunks and MUST stop its + * T-4 timer. + */ + case SCTP_CID_ASCONF: + if (asoc->peer.asconf_capable == 0) + break; + + asoc->peer.asconf_capable = 0; + sctp_add_cmd_sf(cmds, SCTP_CMD_TIMER_STOP, SCTP_TO(SCTP_EVENT_TIMEOUT_T4_RTO)); + break; + default: + break; + } break; + } default: break; } - break; - } - default: - break; } } diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c index 3a0cd075914f..f88dfded0e3a 100644 --- a/net/sctp/sm_statefuns.c +++ b/net/sctp/sm_statefuns.c @@ -3163,7 +3163,6 @@ sctp_disposition_t sctp_sf_operr_notify(const struct sctp_endpoint *ep, sctp_cmd_seq_t *commands) { struct sctp_chunk *chunk = arg; - struct sctp_ulpevent *ev; if (!sctp_vtag_verify(chunk, asoc)) return sctp_sf_pdiscard(ep, asoc, type, arg, commands); @@ -3173,21 +3172,10 @@ sctp_disposition_t sctp_sf_operr_notify(const struct sctp_endpoint *ep, return sctp_sf_violation_chunklen(ep, asoc, type, arg, commands); - while (chunk->chunk_end > chunk->skb->data) { - ev = sctp_ulpevent_make_remote_error(asoc, chunk, 0, - GFP_ATOMIC); - if (!ev) - goto nomem; + sctp_add_cmd_sf(commands, SCTP_CMD_PROCESS_OPERR, + SCTP_CHUNK(chunk)); - sctp_add_cmd_sf(commands, SCTP_CMD_EVENT_ULP, - SCTP_ULPEVENT(ev)); - sctp_add_cmd_sf(commands, SCTP_CMD_PROCESS_OPERR, - SCTP_CHUNK(chunk)); - } return SCTP_DISPOSITION_CONSUME; - -nomem: - return SCTP_DISPOSITION_NOMEM; } /* From 07555c9880da3e2e96e5eae00a03b44cc076deaf Mon Sep 17 00:00:00 2001 From: Russell King Date: Mon, 2 Mar 2009 22:29:37 -0800 Subject: [PATCH 62/92] OMAP: enable smc911x support for LDP platform The following patch enables SMC911x support to work on the OMAP LDP board. Although the SMC911x driver will eventually be obsoleted, the smsc911x patches are rather invasive for the -rc kernels. Rather than risk destablising smsc911x, this simpler patch is preferred to allow the network interface to work. Signed-off-by: Russell King Acked-by: Tony Lindgren Signed-off-by: David S. Miller --- arch/arm/mach-omap2/board-ldp.c | 2 +- drivers/net/smc911x.h | 12 ++++++++++++ 2 files changed, 13 insertions(+), 1 deletion(-) diff --git a/arch/arm/mach-omap2/board-ldp.c b/arch/arm/mach-omap2/board-ldp.c index f6a13451d1fd..6031e179926b 100644 --- a/arch/arm/mach-omap2/board-ldp.c +++ b/arch/arm/mach-omap2/board-ldp.c @@ -81,7 +81,7 @@ static inline void __init ldp_init_smc911x(void) } ldp_smc911x_resources[0].start = cs_mem_base + 0x0; - ldp_smc911x_resources[0].end = cs_mem_base + 0xf; + ldp_smc911x_resources[0].end = cs_mem_base + 0xff; udelay(100); eth_gpio = LDP_SMC911X_GPIO; diff --git a/drivers/net/smc911x.h b/drivers/net/smc911x.h index 870b4c33f108..a45952e72018 100644 --- a/drivers/net/smc911x.h +++ b/drivers/net/smc911x.h @@ -42,6 +42,16 @@ #define SMC_USE_16BIT 0 #define SMC_USE_32BIT 1 #define SMC_IRQ_SENSE IRQF_TRIGGER_LOW +#elif defined(CONFIG_ARCH_OMAP34XX) + #define SMC_USE_16BIT 0 + #define SMC_USE_32BIT 1 + #define SMC_IRQ_SENSE IRQF_TRIGGER_LOW + #define SMC_MEM_RESERVED 1 +#elif defined(CONFIG_ARCH_OMAP24XX) + #define SMC_USE_16BIT 0 + #define SMC_USE_32BIT 1 + #define SMC_IRQ_SENSE IRQF_TRIGGER_LOW + #define SMC_MEM_RESERVED 1 #else /* * Default configuration @@ -675,6 +685,7 @@ smc_pxa_dma_outsl(struct smc911x_local *lp, u_long physaddr, #define CHIP_9116 0x0116 #define CHIP_9117 0x0117 #define CHIP_9118 0x0118 +#define CHIP_9211 0x9211 #define CHIP_9215 0x115A #define CHIP_9217 0x117A #define CHIP_9218 0x118A @@ -689,6 +700,7 @@ static const struct chip_id chip_ids[] = { { CHIP_9116, "LAN9116" }, { CHIP_9117, "LAN9117" }, { CHIP_9118, "LAN9118" }, + { CHIP_9211, "LAN9211" }, { CHIP_9215, "LAN9215" }, { CHIP_9217, "LAN9217" }, { CHIP_9218, "LAN9218" }, From 0fc59d3a01820765e5f3a723733728758b0cf577 Mon Sep 17 00:00:00 2001 From: Yinghai Lu Date: Mon, 2 Mar 2009 23:36:13 -0800 Subject: [PATCH 63/92] x86: fix init_memory_mapping() to handle small ranges Impact: fix failed EFI bootup in certain circumstances Ying Huang found init_memory_mapping() has problem with small ranges less than 2M when he tried to direct map the EFI runtime code out of max_low_pfn_mapped. It turns out we never considered that case and didn't check the range... Reported-by: Ying Huang Signed-off-by: Yinghai Lu Cc: Brian Maly LKML-Reference: <49ACDDED.1060508@kernel.org> Signed-off-by: Ingo Molnar --- arch/x86/mm/init_64.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index e6d36b490250..b1352250096e 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -714,6 +714,8 @@ unsigned long __init_refok init_memory_mapping(unsigned long start, pos = start_pfn << PAGE_SHIFT; end_pfn = ((pos + (PMD_SIZE - 1)) >> PMD_SHIFT) << (PMD_SHIFT - PAGE_SHIFT); + if (end_pfn > (end >> PAGE_SHIFT)) + end_pfn = end >> PAGE_SHIFT; if (start_pfn < end_pfn) { nr_range = save_mr(mr, nr_range, start_pfn, end_pfn, 0); pos = end_pfn << PAGE_SHIFT; From 5a5990d3090b03745a9548a6f5edef02095675cf Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Thu, 26 Feb 2009 06:49:24 +0000 Subject: [PATCH 64/92] net: Avoid race between network down and sysfs Signed-off-by: Stephen Hemminger Acked-by: "Eric W. Biederman" Signed-off-by: David S. Miller --- net/core/net-sysfs.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c index 6ac29a46e23e..484f58750eba 100644 --- a/net/core/net-sysfs.c +++ b/net/core/net-sysfs.c @@ -77,7 +77,9 @@ static ssize_t netdev_store(struct device *dev, struct device_attribute *attr, if (endp == buf) goto err; - rtnl_lock(); + if (!rtnl_trylock()) + return -ERESTARTSYS; + if (dev_isalive(net)) { if ((ret = (*set)(net, new)) == 0) ret = len; From b325fddb7f869e6c95a88dc6573220f162e5b89f Mon Sep 17 00:00:00 2001 From: Stephen Hemminger Date: Thu, 26 Feb 2009 06:55:31 +0000 Subject: [PATCH 65/92] ipv6: Fix sysctl unregistration deadlock Signed-off-by: David S. Miller --- net/ipv6/addrconf.c | 14 ++++++++------ 1 file changed, 8 insertions(+), 6 deletions(-) diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c index f9afb452249c..40fabdee42ae 100644 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@ -493,15 +493,17 @@ static void addrconf_forward_change(struct net *net, __s32 newf) read_unlock(&dev_base_lock); } -static void addrconf_fixup_forwarding(struct ctl_table *table, int *p, int old) +static int addrconf_fixup_forwarding(struct ctl_table *table, int *p, int old) { struct net *net; net = (struct net *)table->extra2; if (p == &net->ipv6.devconf_dflt->forwarding) - return; + return 0; + + if (!rtnl_trylock()) + return -ERESTARTSYS; - rtnl_lock(); if (p == &net->ipv6.devconf_all->forwarding) { __s32 newf = net->ipv6.devconf_all->forwarding; net->ipv6.devconf_dflt->forwarding = newf; @@ -512,6 +514,7 @@ static void addrconf_fixup_forwarding(struct ctl_table *table, int *p, int old) if (*p) rt6_purge_dflt_routers(net); + return 1; } #endif @@ -3983,7 +3986,7 @@ int addrconf_sysctl_forward(ctl_table *ctl, int write, struct file * filp, ret = proc_dointvec(ctl, write, filp, buffer, lenp, ppos); if (write) - addrconf_fixup_forwarding(ctl, valp, val); + ret = addrconf_fixup_forwarding(ctl, valp, val); return ret; } @@ -4019,8 +4022,7 @@ static int addrconf_sysctl_forward_strategy(ctl_table *table, } *valp = new; - addrconf_fixup_forwarding(table, valp, val); - return 1; + return addrconf_fixup_forwarding(table, valp, val); } static struct addrconf_sysctl_table From 176c39af29bc4edaf37f663553eeaacd47b5bc9c Mon Sep 17 00:00:00 2001 From: Daniel Lezcano Date: Tue, 3 Mar 2009 01:06:45 -0800 Subject: [PATCH 66/92] netns: fix addrconf_ifdown kernel panic When a network namespace is destroyed the network interfaces are all unregistered, making addrconf_ifdown called by the netdevice notifier. In the other hand, the addrconf exit method does a loop on the network devices and does addrconf_ifdown on each of them. But the ordering of the netns subsystem is not right because it uses the register_pernet_device instead of register_pernet_subsys. If we handle the loopback as any network device, we can safely use register_pernet_subsys. But if we use register_pernet_subsys, the addrconf exit method will do exactly what was already done with the unregistering of the network devices. So in definitive, this code is pointless. I removed the netns addrconf exit method and moved the code to the addrconf cleanup function. Signed-off-by: Daniel Lezcano Signed-off-by: David S. Miller --- net/ipv6/addrconf.c | 39 +++++++++------------------------------ 1 file changed, 9 insertions(+), 30 deletions(-) diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c index 40fabdee42ae..1220e2c7831e 100644 --- a/net/ipv6/addrconf.c +++ b/net/ipv6/addrconf.c @@ -2611,9 +2611,6 @@ static int addrconf_ifdown(struct net_device *dev, int how) ASSERT_RTNL(); - if ((dev->flags & IFF_LOOPBACK) && how == 1) - how = 0; - rt6_ifdown(net, dev); neigh_ifdown(&nd_tbl, dev); @@ -4448,25 +4445,6 @@ int unregister_inet6addr_notifier(struct notifier_block *nb) EXPORT_SYMBOL(unregister_inet6addr_notifier); -static void addrconf_net_exit(struct net *net) -{ - struct net_device *dev; - - rtnl_lock(); - /* clean dev list */ - for_each_netdev(net, dev) { - if (__in6_dev_get(dev) == NULL) - continue; - addrconf_ifdown(dev, 1); - } - addrconf_ifdown(net->loopback_dev, 2); - rtnl_unlock(); -} - -static struct pernet_operations addrconf_net_ops = { - .exit = addrconf_net_exit, -}; - /* * Init / cleanup code */ @@ -4508,10 +4486,6 @@ int __init addrconf_init(void) if (err) goto errlo; - err = register_pernet_device(&addrconf_net_ops); - if (err) - return err; - register_netdevice_notifier(&ipv6_dev_notf); addrconf_verify(0); @@ -4541,15 +4515,22 @@ errlo: void addrconf_cleanup(void) { struct inet6_ifaddr *ifa; + struct net_device *dev; int i; unregister_netdevice_notifier(&ipv6_dev_notf); - unregister_pernet_device(&addrconf_net_ops); - unregister_pernet_subsys(&addrconf_ops); rtnl_lock(); + /* clean dev list */ + for_each_netdev(&init_net, dev) { + if (__in6_dev_get(dev) == NULL) + continue; + addrconf_ifdown(dev, 1); + } + addrconf_ifdown(init_net.loopback_dev, 2); + /* * Check hash table. */ @@ -4570,6 +4551,4 @@ void addrconf_cleanup(void) del_timer(&addr_chk_timer); rtnl_unlock(); - - unregister_pernet_subsys(&addrconf_net_ops); } From 6eb0777228f31932fc941eafe8b08848466630a1 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sun, 22 Feb 2009 00:09:14 -0800 Subject: [PATCH 67/92] netns: Fix icmp shutdown. Recently I had a kernel panic in icmp_send during a network namespace cleanup. There were packets in the arp queue that failed to be sent and we attempted to generate an ICMP host unreachable message, but failed because icmp_sk_exit had already been called. The network devices are removed from a network namespace and their arp queues are flushed before we do attempt to shutdown subsystems so this error should have been impossible. It turns out icmp_init is using register_pernet_device instead of register_pernet_subsys. Which resulted in icmp being shut down while we still had the possibility of packets in flight, making a nasty NULL pointer deference in interrupt context possible. Changing this to register_pernet_subsys fixes the problem in my testing. Signed-off-by: Eric W. Biederman Acked-by: Denis V. Lunev Signed-off-by: David S. Miller --- net/ipv4/icmp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv4/icmp.c b/net/ipv4/icmp.c index 705b33b184a3..fc562d29cc46 100644 --- a/net/ipv4/icmp.c +++ b/net/ipv4/icmp.c @@ -1205,7 +1205,7 @@ static struct pernet_operations __net_initdata icmp_sk_ops = { int __init icmp_init(void) { - return register_pernet_device(&icmp_sk_ops); + return register_pernet_subsys(&icmp_sk_ops); } EXPORT_SYMBOL(icmp_err_convert); From 2f20d2e667ab1ca44cde5fb361386dff5bb6081d Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sun, 22 Feb 2009 00:10:18 -0800 Subject: [PATCH 68/92] tcp: Like icmp use register_pernet_subsys To remove the possibility of packets flying around when network devices are being cleaned up use reisger_pernet_subsys instead of register_pernet_device. Signed-off-by: Eric W. Biederman Acked-by: Denis V. Lunev Signed-off-by: David S. Miller --- net/ipv4/tcp_ipv4.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c index 19d7b429a262..cf74c416831a 100644 --- a/net/ipv4/tcp_ipv4.c +++ b/net/ipv4/tcp_ipv4.c @@ -2443,7 +2443,7 @@ static struct pernet_operations __net_initdata tcp_sk_ops = { void __init tcp_v4_init(void) { inet_hashinfo_init(&tcp_hashinfo); - if (register_pernet_device(&tcp_sk_ops)) + if (register_pernet_subsys(&tcp_sk_ops)) panic("Failed to create the TCP control socket.\n"); } From 17edde520927070a6bf14a6a75027c0b843443e5 Mon Sep 17 00:00:00 2001 From: "Eric W. Biederman" Date: Sun, 22 Feb 2009 00:11:09 -0800 Subject: [PATCH 69/92] netns: Remove net_alive It turns out that net_alive is unnecessary, and the original problem that led to it being added was simply that the icmp code thought it was a network device and wound up being unable to handle packets while there were still packets in the network namespace. Now that icmp and tcp have been fixed to properly register themselves this problem is no longer present and we have a stronger guarantee that packets will not arrive in a network namespace then that provided by net_alive in netif_receive_skb. So remove net_alive allowing packet reception run a little faster. Additionally document the strong reason why network namespace cleanup is safe so that if something happens again someone else will have a chance of figuring it out. Signed-off-by: Eric W. Biederman Signed-off-by: David S. Miller --- include/net/net_namespace.h | 27 +++++++++++++++++---------- net/core/dev.c | 6 ------ net/core/net_namespace.c | 3 --- 3 files changed, 17 insertions(+), 19 deletions(-) diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 6fc13d905c5f..ded434b032a4 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h @@ -109,11 +109,6 @@ extern struct list_head net_namespace_list; #ifdef CONFIG_NET_NS extern void __put_net(struct net *net); -static inline int net_alive(struct net *net) -{ - return net && atomic_read(&net->count); -} - static inline struct net *get_net(struct net *net) { atomic_inc(&net->count); @@ -145,11 +140,6 @@ int net_eq(const struct net *net1, const struct net *net2) } #else -static inline int net_alive(struct net *net) -{ - return 1; -} - static inline struct net *get_net(struct net *net) { return net; @@ -234,6 +224,23 @@ struct pernet_operations { void (*exit)(struct net *net); }; +/* + * Use these carefully. If you implement a network device and it + * needs per network namespace operations use device pernet operations, + * otherwise use pernet subsys operations. + * + * This is critically important. Most of the network code cleanup + * runs with the assumption that dev_remove_pack has been called so no + * new packets will arrive during and after the cleanup functions have + * been called. dev_remove_pack is not per namespace so instead the + * guarantee of no more packets arriving in a network namespace is + * provided by ensuring that all network devices and all sockets have + * left the network namespace before the cleanup methods are called. + * + * For the longest time the ipv4 icmp code was registered as a pernet + * device which caused kernel oops, and panics during network + * namespace cleanup. So please don't get this wrong. + */ extern int register_pernet_subsys(struct pernet_operations *); extern void unregister_pernet_subsys(struct pernet_operations *); extern int register_pernet_gen_subsys(int *id, struct pernet_operations *); diff --git a/net/core/dev.c b/net/core/dev.c index 72b0d26fd46d..9e4afe650e7a 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -2267,12 +2267,6 @@ int netif_receive_skb(struct sk_buff *skb) rcu_read_lock(); - /* Don't receive packets in an exiting network namespace */ - if (!net_alive(dev_net(skb->dev))) { - kfree_skb(skb); - goto out; - } - #ifdef CONFIG_NET_CLS_ACT if (skb->tc_verd & TC_NCLS) { skb->tc_verd = CLR_TC_NCLS(skb->tc_verd); diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index 2adb1a7d361f..e3bebd36f053 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -157,9 +157,6 @@ static void cleanup_net(struct work_struct *work) struct pernet_operations *ops; struct net *net; - /* Be very certain incoming network packets will not find us */ - rcu_barrier(); - net = container_of(work, struct net, work); mutex_lock(&net_mutex); From 780eef9492b16a1543a3b2ae9f9526a735fc9856 Mon Sep 17 00:00:00 2001 From: Tim Blechmann Date: Thu, 19 Feb 2009 17:34:03 +0100 Subject: [PATCH 70/92] x86: oprofile: don't set counter width from cpuid on Core2 Impact: fix stuck NMIs and non-working oprofile on certain CPUs Resetting the counter width of the performance counters on Intel's Core2 CPUs, breaks the delivery of NMIs, when running in x86_64 mode. This should fix bug #12395: http://bugzilla.kernel.org/show_bug.cgi?id=12395 Signed-off-by: Tim Blechmann Signed-off-by: Robert Richter LKML-Reference: <20090303100412.GC10085@erda.amd.com> Cc: Signed-off-by: Ingo Molnar --- arch/x86/oprofile/op_model_ppro.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/arch/x86/oprofile/op_model_ppro.c b/arch/x86/oprofile/op_model_ppro.c index e9f80c744cf3..10131fbdaada 100644 --- a/arch/x86/oprofile/op_model_ppro.c +++ b/arch/x86/oprofile/op_model_ppro.c @@ -78,8 +78,18 @@ static void ppro_setup_ctrs(struct op_msrs const * const msrs) if (cpu_has_arch_perfmon) { union cpuid10_eax eax; eax.full = cpuid_eax(0xa); - if (counter_width < eax.split.bit_width) - counter_width = eax.split.bit_width; + + /* + * For Core2 (family 6, model 15), don't reset the + * counter width: + */ + if (!(eax.split.version_id == 0 && + current_cpu_data.x86 == 6 && + current_cpu_data.x86_model == 15)) { + + if (counter_width < eax.split.bit_width) + counter_width = eax.split.bit_width; + } } /* clear all counters */ From 25ef4a67e78e1322d55f0a38783537ed89addc02 Mon Sep 17 00:00:00 2001 From: Seth Forshee Date: Mon, 2 Mar 2009 22:39:36 +0100 Subject: [PATCH 71/92] [ARM] 5416/1: Use unused address in v6_early_abort The target of the strex instruction to clear the exlusive monitor is currently the top of the stack. If the store succeeeds this corrupts r0 in pt_regs. Use the next stack location instead of the current one to prevent any chance of corrupting an in-use address. Signed-off-by: Seth Forshee Signed-off-by: Russell King --- arch/arm/mm/abort-ev6.S | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/arch/arm/mm/abort-ev6.S b/arch/arm/mm/abort-ev6.S index 8a7f65ba14b7..94077fbd96b7 100644 --- a/arch/arm/mm/abort-ev6.S +++ b/arch/arm/mm/abort-ev6.S @@ -23,7 +23,8 @@ ENTRY(v6_early_abort) #ifdef CONFIG_CPU_32v6K clrex #else - strex r0, r1, [sp] @ Clear the exclusive monitor + sub r1, sp, #4 @ Get unused stack location + strex r0, r1, [r1] @ Clear the exclusive monitor #endif mrc p15, 0, r1, c5, c0, 0 @ get FSR mrc p15, 0, r0, c6, c0, 0 @ get FAR From b57ee99fab25dbc12150fe66fe54dc52bc6de784 Mon Sep 17 00:00:00 2001 From: Catalin Marinas Date: Tue, 3 Mar 2009 11:44:12 +0100 Subject: [PATCH 72/92] [ARM] 5417/1: Set the correct cacheid for ARMv6 CPUs with ARMv7 style MMU The cacheid_init() function assumes that if cpu_architecture() returns 7, the caches are VIPT_NONALIASING. The cpu_architecture() function returns the version of the supported MMU features (e.g. TEX remapping) but it doesn't make any assumptions about the cache type. The patch adds the checking of the Cache Type Register for the ARMv7 format. Signed-off-by: Catalin Marinas Signed-off-by: Russell King --- arch/arm/kernel/setup.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-) diff --git a/arch/arm/kernel/setup.c b/arch/arm/kernel/setup.c index 7049815d66d5..68d6494c0389 100644 --- a/arch/arm/kernel/setup.c +++ b/arch/arm/kernel/setup.c @@ -233,12 +233,13 @@ static void __init cacheid_init(void) unsigned int cachetype = read_cpuid_cachetype(); unsigned int arch = cpu_architecture(); - if (arch >= CPU_ARCH_ARMv7) { - cacheid = CACHEID_VIPT_NONALIASING; - if ((cachetype & (3 << 14)) == 1 << 14) - cacheid |= CACHEID_ASID_TAGGED; - } else if (arch >= CPU_ARCH_ARMv6) { - if (cachetype & (1 << 23)) + if (arch >= CPU_ARCH_ARMv6) { + if ((cachetype & (7 << 29)) == 4 << 29) { + /* ARMv7 register format */ + cacheid = CACHEID_VIPT_NONALIASING; + if ((cachetype & (3 << 14)) == 1 << 14) + cacheid |= CACHEID_ASID_TAGGED; + } else if (cachetype & (1 << 23)) cacheid = CACHEID_VIPT_ALIASING; else cacheid = CACHEID_VIPT_NONALIASING; From bdf602bd737eb07d63d6fa2da826b4751fdf9bab Mon Sep 17 00:00:00 2001 From: Russell King Date: Tue, 3 Mar 2009 13:43:47 +0000 Subject: [PATCH 73/92] [ARM] fix lots of ARM __devexit sillyness MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit `iop_adma_remove' referenced in section `.data' of drivers/built-in.o: defined in discarded section `.devexit.text' of drivers/built-in.o `mv_xor_remove' referenced in section `.data' of drivers/built-in.o: defined in discarded section `.devexit.text' of drivers/built-in.o `mv64xxx_i2c_unmap_regs' referenced in section `.devinit.text' of drivers/built-in.o: defined in discarded section `.devexit.text' of drivers/built-in.o `mv64xxx_i2c_remove' referenced in section `.data' of drivers/built-in.o: defined in discarded section `.devexit.text' of drivers/built-in.o `orion_nand_remove' referenced in section `.data' of drivers/built-in.o: defined in discarded section `.devexit.text' of drivers/built-in.o `pxafb_remove' referenced in section `.data' of drivers/built-in.o: defined in discarded section `.devexit.text' of drivers/built-in.o Acked-by: Uwe Kleine-König Signed-off-by: Russell King --- drivers/dma/iop-adma.c | 2 +- drivers/dma/mv_xor.c | 2 +- drivers/i2c/busses/i2c-mv64xxx.c | 4 ++-- drivers/mtd/nand/orion_nand.c | 2 +- drivers/video/pxafb.c | 2 +- 5 files changed, 6 insertions(+), 6 deletions(-) diff --git a/drivers/dma/iop-adma.c b/drivers/dma/iop-adma.c index ea5440dd10dc..647374acba94 100644 --- a/drivers/dma/iop-adma.c +++ b/drivers/dma/iop-adma.c @@ -1401,7 +1401,7 @@ MODULE_ALIAS("platform:iop-adma"); static struct platform_driver iop_adma_driver = { .probe = iop_adma_probe, - .remove = iop_adma_remove, + .remove = __devexit_p(iop_adma_remove), .driver = { .owner = THIS_MODULE, .name = "iop-adma", diff --git a/drivers/dma/mv_xor.c b/drivers/dma/mv_xor.c index d35cbd1ff0b3..5d5d5b31867f 100644 --- a/drivers/dma/mv_xor.c +++ b/drivers/dma/mv_xor.c @@ -1287,7 +1287,7 @@ mv_xor_conf_mbus_windows(struct mv_xor_shared_private *msp, static struct platform_driver mv_xor_driver = { .probe = mv_xor_probe, - .remove = mv_xor_remove, + .remove = __devexit_p(mv_xor_remove), .driver = { .owner = THIS_MODULE, .name = MV_XOR_NAME, diff --git a/drivers/i2c/busses/i2c-mv64xxx.c b/drivers/i2c/busses/i2c-mv64xxx.c index eeda276f8f16..7f186bbcb99d 100644 --- a/drivers/i2c/busses/i2c-mv64xxx.c +++ b/drivers/i2c/busses/i2c-mv64xxx.c @@ -482,7 +482,7 @@ mv64xxx_i2c_map_regs(struct platform_device *pd, return 0; } -static void __devexit +static void mv64xxx_i2c_unmap_regs(struct mv64xxx_i2c_data *drv_data) { if (drv_data->reg_base) { @@ -577,7 +577,7 @@ mv64xxx_i2c_remove(struct platform_device *dev) static struct platform_driver mv64xxx_i2c_driver = { .probe = mv64xxx_i2c_probe, - .remove = mv64xxx_i2c_remove, + .remove = __devexit_p(mv64xxx_i2c_remove), .driver = { .owner = THIS_MODULE, .name = MV64XXX_I2C_CTLR_NAME, diff --git a/drivers/mtd/nand/orion_nand.c b/drivers/mtd/nand/orion_nand.c index 917cf8d3ae95..c2dfd3ea353d 100644 --- a/drivers/mtd/nand/orion_nand.c +++ b/drivers/mtd/nand/orion_nand.c @@ -149,7 +149,7 @@ static int __devexit orion_nand_remove(struct platform_device *pdev) static struct platform_driver orion_nand_driver = { .probe = orion_nand_probe, - .remove = orion_nand_remove, + .remove = __devexit_p(orion_nand_remove), .driver = { .name = "orion_nand", .owner = THIS_MODULE, diff --git a/drivers/video/pxafb.c b/drivers/video/pxafb.c index 48ff701d3a72..2552b9f325ee 100644 --- a/drivers/video/pxafb.c +++ b/drivers/video/pxafb.c @@ -2230,7 +2230,7 @@ static int __devexit pxafb_remove(struct platform_device *dev) static struct platform_driver pxafb_driver = { .probe = pxafb_probe, - .remove = pxafb_remove, + .remove = __devexit_p(pxafb_remove), .suspend = pxafb_suspend, .resume = pxafb_resume, .driver = { From 1777f1a978153e8b887c1e1eb5160ac46098b142 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Krzysztof=20Ha=C5=82asa?= Date: Wed, 4 Mar 2009 08:01:22 +0800 Subject: [PATCH 74/92] crypto: ixp4xx - Fix qmgr_request_queue build failure MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit There is another user of IXP4xx queue manager, fix it. Signed-off-by: Krzysztof Hałasa Signed-off-by: Herbert Xu --- drivers/crypto/ixp4xx_crypto.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/crypto/ixp4xx_crypto.c b/drivers/crypto/ixp4xx_crypto.c index 2d637e0fbc03..d9e751be8c5f 100644 --- a/drivers/crypto/ixp4xx_crypto.c +++ b/drivers/crypto/ixp4xx_crypto.c @@ -457,10 +457,12 @@ static int init_ixp_crypto(void) if (!ctx_pool) { goto err; } - ret = qmgr_request_queue(SEND_QID, NPE_QLEN_TOTAL, 0, 0); + ret = qmgr_request_queue(SEND_QID, NPE_QLEN_TOTAL, 0, 0, + "ixp_crypto:out", NULL); if (ret) goto err; - ret = qmgr_request_queue(RECV_QID, NPE_QLEN, 0, 0); + ret = qmgr_request_queue(RECV_QID, NPE_QLEN, 0, 0, + "ixp_crypto:in", NULL); if (ret) { qmgr_release_queue(SEND_QID); goto err; From fec6c6fec3e20637bee5d276fb61dd8b49a3f9cc Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Tue, 3 Mar 2009 17:05:22 -0800 Subject: [PATCH 75/92] Linux 2.6.29-rc7 --- Makefile | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/Makefile b/Makefile index df6ce3e80090..d04ee0ad1dcc 100644 --- a/Makefile +++ b/Makefile @@ -1,7 +1,7 @@ VERSION = 2 PATCHLEVEL = 6 SUBLEVEL = 29 -EXTRAVERSION = -rc6 +EXTRAVERSION = -rc7 NAME = Erotic Pickled Herring # *DOCUMENTATION* From 368a12117dd8abf6eaefa37c21ac313b517128b9 Mon Sep 17 00:00:00 2001 From: Tony Breeds Date: Tue, 3 Mar 2009 17:59:30 +0000 Subject: [PATCH 76/92] powerpc: Run sbc610 USB fixup code only on the appropriate platform. commit a969e76a7101bf5f3d369563df1ca1253dd6131b (powerpc: Correct USB support for GE Fanuc SBC610) introduced a fixup for NEC usb controllers. This fixup should only run on GEF SBC610 boards. Fixes Fedora bug #486511. (https://bugzilla.redhat.com/show_bug.cgi?id=486511) Signed-off-by: Tony Breeds Signed-off-by: Benjamin Herrenschmidt --- arch/powerpc/platforms/86xx/gef_sbc610.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/arch/powerpc/platforms/86xx/gef_sbc610.c b/arch/powerpc/platforms/86xx/gef_sbc610.c index fb371f5ce132..d6b772ba3b8f 100644 --- a/arch/powerpc/platforms/86xx/gef_sbc610.c +++ b/arch/powerpc/platforms/86xx/gef_sbc610.c @@ -142,6 +142,10 @@ static void __init gef_sbc610_nec_fixup(struct pci_dev *pdev) { unsigned int val; + /* Do not do the fixup on other platforms! */ + if (!machine_is(gef_sbc610)) + return; + printk(KERN_INFO "Running NEC uPD720101 Fixup\n"); /* Ensure ports 1, 2, 3, 4 & 5 are enabled */ From 4843b93c96ae5043c6279c4ec6fcd8ee3866ff5b Mon Sep 17 00:00:00 2001 From: Pablo Neira Ayuso Date: Tue, 3 Mar 2009 23:37:30 -0800 Subject: [PATCH 77/92] netlink: invert error code in netlink_set_err() The callers of netlink_set_err() currently pass a negative value as parameter for the error code. However, sk->sk_err wants a positive error value. Without this patch, skb_recv_datagram() called by netlink_recvmsg() may return a positive value to report an error. Another choice to fix this is to change callers to pass a positive error value, but this seems a bit inconsistent and error prone to me. Indeed, the callers of netlink_set_err() assumed that the (usual) negative value for error codes was fine before this patch :). This patch also includes some documentation in docbook format for netlink_set_err() to avoid this sort of confusion. Signed-off-by: Pablo Neira Ayuso Signed-off-by: David S. Miller --- net/netlink/af_netlink.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/net/netlink/af_netlink.c b/net/netlink/af_netlink.c index 9eb895c7a2a9..3ae3cb816563 100644 --- a/net/netlink/af_netlink.c +++ b/net/netlink/af_netlink.c @@ -1084,6 +1084,13 @@ out: return 0; } +/** + * netlink_set_err - report error to broadcast listeners + * @ssk: the kernel netlink socket, as returned by netlink_kernel_create() + * @pid: the PID of a process that we want to skip (if any) + * @groups: the broadcast group that will notice the error + * @code: error code, must be negative (as usual in kernelspace) + */ void netlink_set_err(struct sock *ssk, u32 pid, u32 group, int code) { struct netlink_set_err_data info; @@ -1093,7 +1100,8 @@ void netlink_set_err(struct sock *ssk, u32 pid, u32 group, int code) info.exclude_sk = ssk; info.pid = pid; info.group = group; - info.code = code; + /* sk->sk_err wants a positive error value */ + info.code = -code; read_lock(&nl_table_lock); From a1a69c8db7f988f903349442a7538d21b56c38e9 Mon Sep 17 00:00:00 2001 From: Peter Korsgaard Date: Tue, 3 Mar 2009 23:48:16 -0800 Subject: [PATCH 78/92] dm9601: new vendor/product IDs Add vendor/product IDs for new no name dm9601 compatible usb ethernet adaptors. Reported-by: Eric Lauriault Signed-off-by: Peter Korsgaard Signed-off-by: David S. Miller --- drivers/net/usb/dm9601.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/drivers/net/usb/dm9601.c b/drivers/net/usb/dm9601.c index 5b67bbf1987e..81682c6defa0 100644 --- a/drivers/net/usb/dm9601.c +++ b/drivers/net/usb/dm9601.c @@ -633,6 +633,10 @@ static const struct usb_device_id products[] = { }, { USB_DEVICE(0x0a47, 0x9601), /* Hirose USB-100 */ + .driver_info = (unsigned long)&dm9601_info, + }, + { + USB_DEVICE(0x0fe6, 0x8101), /* DM9601 USB to Fast Ethernet Adapter */ .driver_info = (unsigned long)&dm9601_info, }, {}, // END From 4222474519ff5b31a526dfa1da7aa4b0e38bef5c Mon Sep 17 00:00:00 2001 From: Meelis Roos Date: Tue, 3 Mar 2009 23:48:50 -0800 Subject: [PATCH 79/92] net: fix tokenring license Currently, modular tokenring ("tr") lacks a license and fails to load: tr: module license 'unspecified' taints kernel. tr: Unknown symbol proc_net_fops_create Beacuse of this, no tokenring driver can load if it depends on modular tr. Fix this by adding GPL module license as it is in the kernel. With this fix, tr module loads fine and tms380 driver also loads. Well, it does'nt work but that's a different bug. Signed-off-by: Meelis Roos Signed-off-by: David S. Miller --- net/802/tr.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/net/802/tr.c b/net/802/tr.c index 158150fee462..f47ae289d83b 100644 --- a/net/802/tr.c +++ b/net/802/tr.c @@ -668,3 +668,5 @@ module_init(rif_init); EXPORT_SYMBOL(tr_type_trans); EXPORT_SYMBOL(alloc_trdev); + +MODULE_LICENSE("GPL"); From 858b9ced6e73a0f087294c398a1ae70a7eeed94f Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Wed, 4 Mar 2009 00:11:42 -0800 Subject: [PATCH 80/92] net: more timeouts that reach -1 with while (timeout-- > 0); timeout reaches -1 after the loop, so the tests below are off by one. also don't do an '< 0' test on an unsigned. Signed-off-by: Roel Kluin Signed-off-by: David S. Miller --- drivers/net/arm/ks8695net.c | 2 +- drivers/net/jme.c | 3 ++- drivers/net/ucc_geth_mii.c | 4 ++-- 3 files changed, 5 insertions(+), 4 deletions(-) diff --git a/drivers/net/arm/ks8695net.c b/drivers/net/arm/ks8695net.c index 1cf2f949c0b4..f3a127434897 100644 --- a/drivers/net/arm/ks8695net.c +++ b/drivers/net/arm/ks8695net.c @@ -560,7 +560,7 @@ ks8695_reset(struct ks8695_priv *ksp) msleep(1); } - if (reset_timeout == 0) { + if (reset_timeout < 0) { dev_crit(ksp->dev, "Timeout waiting for DMA engines to reset\n"); /* And blithely carry on */ diff --git a/drivers/net/jme.c b/drivers/net/jme.c index 08b34051c646..a6e1a35a13cb 100644 --- a/drivers/net/jme.c +++ b/drivers/net/jme.c @@ -957,13 +957,14 @@ jme_process_receive(struct jme_adapter *jme, int limit) goto out_inc; i = atomic_read(&rxring->next_to_clean); - while (limit-- > 0) { + while (limit > 0) { rxdesc = rxring->desc; rxdesc += i; if ((rxdesc->descwb.flags & cpu_to_le16(RXWBFLAG_OWN)) || !(rxdesc->descwb.desccnt & RXWBDCNT_WBCPL)) goto out; + --limit; desccnt = rxdesc->descwb.desccnt & RXWBDCNT_DCNT; diff --git a/drivers/net/ucc_geth_mii.c b/drivers/net/ucc_geth_mii.c index 54635911305c..0ada4edd56eb 100644 --- a/drivers/net/ucc_geth_mii.c +++ b/drivers/net/ucc_geth_mii.c @@ -107,7 +107,7 @@ int uec_mdio_read(struct mii_bus *bus, int mii_id, int regnum) static int uec_mdio_reset(struct mii_bus *bus) { struct ucc_mii_mng __iomem *regs = (void __iomem *)bus->priv; - unsigned int timeout = PHY_INIT_TIMEOUT; + int timeout = PHY_INIT_TIMEOUT; mutex_lock(&bus->mdio_lock); @@ -123,7 +123,7 @@ static int uec_mdio_reset(struct mii_bus *bus) mutex_unlock(&bus->mdio_lock); - if (timeout <= 0) { + if (timeout < 0) { printk(KERN_ERR "%s: The MII Bus is stuck!\n", bus->name); return -EBUSY; } From b9bdcd9bd78d253dcc8e13c29f0acd67e080e7c1 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Wed, 4 Mar 2009 00:05:56 -0800 Subject: [PATCH 81/92] net pcmcia: worklimit reaches -1 with while (--worklimit >= 0); worklimit reaches -1 after the loop. In 3c589_cs.c this caused a warning not to be printed. In 3c574_cs.c contrastingly, el3_rx() treats worklimit differently: static int el3_rx(struct net_device *dev, int worklimit) { while (--worklimit >= 0) { ... } return worklimit; } el3_rx() is only called by function el3_interrupt(): twice: static irqreturn_t el3_interrupt(int irq, void *dev_id) { int work_budget = max_interrupt_work; while(...) { if (...) work_budget = el3_rx(dev, work_budget); if (...) work_budget = el3_rx(dev, work_budget); if (--work_budget < 0) { ... break; } } } The error path can occur 2 too early. Signed-off-by: Roel Kluin Signed-off-by: David S. Miller --- drivers/net/pcmcia/3c574_cs.c | 3 ++- drivers/net/pcmcia/3c589_cs.c | 3 ++- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/drivers/net/pcmcia/3c574_cs.c b/drivers/net/pcmcia/3c574_cs.c index e5cb6b1f0ebd..2404a838b1fe 100644 --- a/drivers/net/pcmcia/3c574_cs.c +++ b/drivers/net/pcmcia/3c574_cs.c @@ -1035,7 +1035,8 @@ static int el3_rx(struct net_device *dev, int worklimit) DEBUG(3, "%s: in rx_packet(), status %4.4x, rx_status %4.4x.\n", dev->name, inw(ioaddr+EL3_STATUS), inw(ioaddr+RxStatus)); while (!((rx_status = inw(ioaddr + RxStatus)) & 0x8000) && - (--worklimit >= 0)) { + worklimit > 0) { + worklimit--; if (rx_status & 0x4000) { /* Error, update stats. */ short error = rx_status & 0x3800; dev->stats.rx_errors++; diff --git a/drivers/net/pcmcia/3c589_cs.c b/drivers/net/pcmcia/3c589_cs.c index 73ecc657999d..1e01b8a6dbf3 100644 --- a/drivers/net/pcmcia/3c589_cs.c +++ b/drivers/net/pcmcia/3c589_cs.c @@ -857,7 +857,8 @@ static int el3_rx(struct net_device *dev) DEBUG(3, "%s: in rx_packet(), status %4.4x, rx_status %4.4x.\n", dev->name, inw(ioaddr+EL3_STATUS), inw(ioaddr+RX_STATUS)); while (!((rx_status = inw(ioaddr + RX_STATUS)) & 0x8000) && - (--worklimit >= 0)) { + worklimit > 0) { + worklimit--; if (rx_status & 0x4000) { /* Error, update stats. */ short error = rx_status & 0x3800; dev->stats.rx_errors++; From 948731115774c2e5ff7409360f35389459502211 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Wed, 4 Mar 2009 00:07:57 -0800 Subject: [PATCH 82/92] aoe: error printed 1 too early with while (i-- > 0); i reaches -1 after the loop, so the test below is printed one too early: 0 still means success. Signed-off-by: Roel Kluin Signed-off-by: David S. Miller --- drivers/block/aoe/aoedev.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/block/aoe/aoedev.c b/drivers/block/aoe/aoedev.c index cc250577d405..eeea477d9601 100644 --- a/drivers/block/aoe/aoedev.c +++ b/drivers/block/aoe/aoedev.c @@ -173,7 +173,7 @@ skbfree(struct sk_buff *skb) return; while (atomic_read(&skb_shinfo(skb)->dataref) != 1 && i-- > 0) msleep(Sms); - if (i <= 0) { + if (i < 0) { printk(KERN_ERR "aoe: %s holds ref: %s\n", skb->dev ? skb->dev->name : "netif", From 4a8fd2cfdad4d043a1fadba2f3f340945d966825 Mon Sep 17 00:00:00 2001 From: Roel Kluin Date: Wed, 4 Mar 2009 00:08:39 -0800 Subject: [PATCH 83/92] sungem: another error printed one too early Another error was printed one too early. Signed-off-by: Roel Kluin Signed-off-by: David S. Miller --- drivers/net/sungem.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/sungem.c b/drivers/net/sungem.c index 8d64b1da0465..0fcb7503363d 100644 --- a/drivers/net/sungem.c +++ b/drivers/net/sungem.c @@ -1229,7 +1229,7 @@ static void gem_reset(struct gem *gp) break; } while (val & (GREG_SWRST_TXRST | GREG_SWRST_RXRST)); - if (limit <= 0) + if (limit < 0) printk(KERN_ERR "%s: SW reset is ghetto.\n", gp->dev->name); if (gp->phy_type == phy_serialink || gp->phy_type == phy_serdes) From fe7ca2e1e847b65c12d245cbf402af89da96888a Mon Sep 17 00:00:00 2001 From: Brian Haley Date: Wed, 4 Mar 2009 03:18:11 -0800 Subject: [PATCH 84/92] IPv6: add "disable" module parameter support to ipv6.ko Add "disable" module parameter support to ipv6.ko by specifying "disable=1" on module load. We just do the minimum of initializing inetsw6[] so calls from other modules to inet6_register_protosw() won't OOPs, then bail out. No IPv6 addresses or sockets can be created as a result, and a reboot is required to enable IPv6. Signed-off-by: Brian Haley Signed-off-by: David S. Miller --- Documentation/networking/ipv6.txt | 35 +++++++++++++++++++++++++++++++ net/ipv6/af_inet6.c | 21 ++++++++++++++----- 2 files changed, 51 insertions(+), 5 deletions(-) create mode 100644 Documentation/networking/ipv6.txt diff --git a/Documentation/networking/ipv6.txt b/Documentation/networking/ipv6.txt new file mode 100644 index 000000000000..268e5c103dd8 --- /dev/null +++ b/Documentation/networking/ipv6.txt @@ -0,0 +1,35 @@ + +Options for the ipv6 module are supplied as parameters at load time. + +Module options may be given as command line arguments to the insmod +or modprobe command, but are usually specified in either the +/etc/modules.conf or /etc/modprobe.conf configuration file, or in a +distro-specific configuration file. + +The available ipv6 module parameters are listed below. If a parameter +is not specified the default value is used. + +The parameters are as follows: + +disable + + Specifies whether to load the IPv6 module, but disable all + its functionality. This might be used when another module + has a dependency on the IPv6 module being loaded, but no + IPv6 addresses or operations are desired. + + The possible values and their effects are: + + 0 + IPv6 is enabled. + + This is the default value. + + 1 + IPv6 is disabled. + + No IPv6 addresses will be added to interfaces, and + it will not be possible to open an IPv6 socket. + + A reboot is required to enable IPv6. + diff --git a/net/ipv6/af_inet6.c b/net/ipv6/af_inet6.c index c802bc1658a8..da944eca2ca6 100644 --- a/net/ipv6/af_inet6.c +++ b/net/ipv6/af_inet6.c @@ -72,6 +72,10 @@ MODULE_LICENSE("GPL"); static struct list_head inetsw6[SOCK_MAX]; static DEFINE_SPINLOCK(inetsw6_lock); +static int disable_ipv6 = 0; +module_param_named(disable, disable_ipv6, int, 0); +MODULE_PARM_DESC(disable, "Disable IPv6 such that it is non-functional"); + static __inline__ struct ipv6_pinfo *inet6_sk_generic(struct sock *sk) { const int offset = sk->sk_prot->obj_size - sizeof(struct ipv6_pinfo); @@ -991,10 +995,21 @@ static int __init inet6_init(void) { struct sk_buff *dummy_skb; struct list_head *r; - int err; + int err = 0; BUILD_BUG_ON(sizeof(struct inet6_skb_parm) > sizeof(dummy_skb->cb)); + /* Register the socket-side information for inet6_create. */ + for(r = &inetsw6[0]; r < &inetsw6[SOCK_MAX]; ++r) + INIT_LIST_HEAD(r); + + if (disable_ipv6) { + printk(KERN_INFO + "IPv6: Loaded, but administratively disabled, " + "reboot required to enable\n"); + goto out; + } + err = proto_register(&tcpv6_prot, 1); if (err) goto out; @@ -1012,10 +1027,6 @@ static int __init inet6_init(void) goto out_unregister_udplite_proto; - /* Register the socket-side information for inet6_create. */ - for(r = &inetsw6[0]; r < &inetsw6[SOCK_MAX]; ++r) - INIT_LIST_HEAD(r); - /* We MUST register RAW sockets before we create the ICMP6, * IGMP6, or NDISC control sockets. */ From fb13d9f9e450bceafd88ac8a98f7a98e8096a5fe Mon Sep 17 00:00:00 2001 From: Brian Haley Date: Wed, 4 Mar 2009 03:20:26 -0800 Subject: [PATCH 85/92] SCTP: change sctp_ctl_sock_init() to try IPv4 if IPv6 fails Change sctp_ctl_sock_init() to try IPv4 if IPv6 socket registration fails. Required if the IPv6 module is loaded with "disable=1", else SCTP will fail to load. Signed-off-by: Brian Haley Signed-off-by: Vlad Yasevich Signed-off-by: David S. Miller --- net/sctp/protocol.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/net/sctp/protocol.c b/net/sctp/protocol.c index 4e6638449639..c4986d0f7419 100644 --- a/net/sctp/protocol.c +++ b/net/sctp/protocol.c @@ -717,15 +717,20 @@ static int sctp_inetaddr_event(struct notifier_block *this, unsigned long ev, static int sctp_ctl_sock_init(void) { int err; - sa_family_t family; + sa_family_t family = PF_INET; if (sctp_get_pf_specific(PF_INET6)) family = PF_INET6; - else - family = PF_INET; err = inet_ctl_sock_create(&sctp_ctl_sock, family, SOCK_SEQPACKET, IPPROTO_SCTP, &init_net); + + /* If IPv6 socket could not be created, try the IPv4 socket */ + if (err < 0 && family == PF_INET6) + err = inet_ctl_sock_create(&sctp_ctl_sock, AF_INET, + SOCK_SEQPACKET, IPPROTO_SCTP, + &init_net); + if (err < 0) { printk(KERN_ERR "SCTP: Failed to create the SCTP control socket.\n"); From 5ad8b7d12605e88d1e532061699102797fdefe08 Mon Sep 17 00:00:00 2001 From: Helge Bahmann Date: Wed, 4 Mar 2009 21:49:14 +1000 Subject: [PATCH 86/92] drm: fix double lock typo [airlied: you shall not retype patches from other trees half asleep] Signed-of-by: Dave Airlie --- drivers/gpu/drm/drm_stub.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/gpu/drm/drm_stub.c b/drivers/gpu/drm/drm_stub.c index 096e2a37446d..7c8b15b22bf2 100644 --- a/drivers/gpu/drm/drm_stub.c +++ b/drivers/gpu/drm/drm_stub.c @@ -168,7 +168,7 @@ int drm_setmaster_ioctl(struct drm_device *dev, void *data, file_priv->minor->master != file_priv->master) { mutex_lock(&dev->struct_mutex); file_priv->minor->master = drm_master_get(file_priv->master); - mutex_lock(&dev->struct_mutex); + mutex_unlock(&dev->struct_mutex); } return 0; From 9f8ac0b7b063be77f0de7a27fe5e6a0aa2cce58d Mon Sep 17 00:00:00 2001 From: Matt Carlson Date: Wed, 25 Feb 2009 14:21:20 +0000 Subject: [PATCH 87/92] tg3: Fix 5906 link problems Commit 6833c043f9fc03696fde623914c4a0277df2a0bc introduced the phy auto-powerdown capability. While the APD feature only works for 5761 and 5784 asic revisions, the (harmless portion of the) code was applied to all 5705 and newer devices. However, the 5906 phy departs from the usual design. This commit was interfering with the 5906's ability to negotiate link against some switches. This patch corrects the problem. Signed-off-by: Matt Carlson Signed-off-by: Benjamin Li Signed-off-by: Michael Chan Signed-off-by: David S. Miller --- drivers/net/tg3.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c index b080f9493d83..dabdf59f8016 100644 --- a/drivers/net/tg3.c +++ b/drivers/net/tg3.c @@ -1473,7 +1473,8 @@ static void tg3_phy_toggle_apd(struct tg3 *tp, bool enable) { u32 reg; - if (!(tp->tg3_flags2 & TG3_FLG2_5705_PLUS)) + if (!(tp->tg3_flags2 & TG3_FLG2_5705_PLUS) || + GET_ASIC_REV(tp->pci_chip_rev_id) == ASIC_REV_5906) return; reg = MII_TG3_MISC_SHDW_WREN | From a883bf564ea555447a76682bb2d8d4bc92e23e0e Mon Sep 17 00:00:00 2001 From: Jarek Poplawski Date: Wed, 4 Mar 2009 17:38:10 -0800 Subject: [PATCH 88/92] pkt_sched: act_police: Fix a rate estimator test. A commit c1b56878fb68e9c14070939ea4537ad4db79ffae "tc: policing requires a rate estimator" introduced a test which invalidates previously working configs, based on examples from iproute2: doc/actions/actions-general. This is too rigorous: a rate estimator is needed only when police's "avrate" option is used. Reported-by: Joao Correia Diagnosed-by: John Dykstra Signed-off-by: Jarek Poplawski Signed-off-by: David S. Miller --- net/sched/act_police.c | 13 ++++++------- 1 file changed, 6 insertions(+), 7 deletions(-) diff --git a/net/sched/act_police.c b/net/sched/act_police.c index 5c72a116b1a4..f8f047b61245 100644 --- a/net/sched/act_police.c +++ b/net/sched/act_police.c @@ -183,13 +183,6 @@ override: if (R_tab == NULL) goto failure; - if (!est && (ret == ACT_P_CREATED || - !gen_estimator_active(&police->tcf_bstats, - &police->tcf_rate_est))) { - err = -EINVAL; - goto failure; - } - if (parm->peakrate.rate) { P_tab = qdisc_get_rtab(&parm->peakrate, tb[TCA_POLICE_PEAKRATE]); @@ -205,6 +198,12 @@ override: &police->tcf_lock, est); if (err) goto failure_unlock; + } else if (tb[TCA_POLICE_AVRATE] && + (ret == ACT_P_CREATED || + !gen_estimator_active(&police->tcf_bstats, + &police->tcf_rate_est))) { + err = -EINVAL; + goto failure_unlock; } /* No failure allowed after this point */ From 87786945fe4b0e60e8f1db62d5ee8a3cec539a67 Mon Sep 17 00:00:00 2001 From: Meelis Roos Date: Wed, 4 Mar 2009 04:59:41 +0000 Subject: [PATCH 89/92] tmspci: fix request_irq race Currently, tmspci tokenring driver crashes on device initialization because it requests its irq before initializing corresponding data structures. Fix this by moving request_irq call to a safer place. Signed-off-by: Meelis Roos Signed-off-by: David S. Miller --- drivers/net/tokenring/tmspci.c | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) diff --git a/drivers/net/tokenring/tmspci.c b/drivers/net/tokenring/tmspci.c index 5f601773c260..e2150b3c83d9 100644 --- a/drivers/net/tokenring/tmspci.c +++ b/drivers/net/tokenring/tmspci.c @@ -121,11 +121,6 @@ static int __devinit tms_pci_attach(struct pci_dev *pdev, const struct pci_devic goto err_out_trdev; } - ret = request_irq(pdev->irq, tms380tr_interrupt, IRQF_SHARED, - dev->name, dev); - if (ret) - goto err_out_region; - dev->base_addr = pci_ioaddr; dev->irq = pci_irq_line; dev->dma = 0; @@ -142,7 +137,7 @@ static int __devinit tms_pci_attach(struct pci_dev *pdev, const struct pci_devic ret = tmsdev_init(dev, &pdev->dev); if (ret) { printk("%s: unable to get memory for dev->priv.\n", dev->name); - goto err_out_irq; + goto err_out_region; } tp = netdev_priv(dev); @@ -157,6 +152,11 @@ static int __devinit tms_pci_attach(struct pci_dev *pdev, const struct pci_devic tp->tmspriv = cardinfo; + ret = request_irq(pdev->irq, tms380tr_interrupt, IRQF_SHARED, + dev->name, dev); + if (ret) + goto err_out_tmsdev; + dev->open = tms380tr_open; dev->stop = tms380tr_close; pci_set_drvdata(pdev, dev); @@ -164,15 +164,15 @@ static int __devinit tms_pci_attach(struct pci_dev *pdev, const struct pci_devic ret = register_netdev(dev); if (ret) - goto err_out_tmsdev; + goto err_out_irq; return 0; +err_out_irq: + free_irq(pdev->irq, dev); err_out_tmsdev: pci_set_drvdata(pdev, NULL); tmsdev_term(dev); -err_out_irq: - free_irq(pdev->irq, dev); err_out_region: release_region(pci_ioaddr, TMS_PCI_IO_EXTENT); err_out_trdev: From 54acd0efab072cb70e87206329d561b297f93bbb Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Wed, 4 Mar 2009 23:01:02 -0800 Subject: [PATCH 90/92] net: Fix missing dev->neigh_setup in register_netdevice(). Signed-off-by: David S. Miller --- net/core/dev.c | 1 + 1 file changed, 1 insertion(+) diff --git a/net/core/dev.c b/net/core/dev.c index 9e4afe650e7a..2dd484ed3dbb 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -4339,6 +4339,7 @@ int register_netdevice(struct net_device *dev) dev->do_ioctl = ops->ndo_do_ioctl; dev->set_config = ops->ndo_set_config; dev->change_mtu = ops->ndo_change_mtu; + dev->neigh_setup = ops->ndo_neigh_setup; dev->tx_timeout = ops->ndo_tx_timeout; dev->get_stats = ops->ndo_get_stats; dev->vlan_rx_register = ops->ndo_vlan_rx_register; From 9d40bbda599def1e1d155d7f7dca14fe8744bd2b Mon Sep 17 00:00:00 2001 From: "David S. Miller" Date: Wed, 4 Mar 2009 23:46:25 -0800 Subject: [PATCH 91/92] vlan: Fix vlan-in-vlan crashes. As analyzed by Patrick McHardy, vlan needs to reset it's netdev_ops pointer in it's ->init() function but this leaves the compat method pointers stale. Add a netdev_resync_ops() and call it from the vlan code. Any other driver which changes ->netdev_ops after register_netdevice() will need to call this new function after doing so too. With help from Patrick McHardy. Tested-by: Patrick McHardy Signed-off-by: David S. Miller --- include/linux/netdevice.h | 1 + net/8021q/vlan_dev.c | 3 ++- net/core/dev.c | 56 ++++++++++++++++++++++++--------------- 3 files changed, 37 insertions(+), 23 deletions(-) diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index ec54785d34f9..659366734f3f 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -1079,6 +1079,7 @@ extern void synchronize_net(void); extern int register_netdevice_notifier(struct notifier_block *nb); extern int unregister_netdevice_notifier(struct notifier_block *nb); extern int init_dummy_netdev(struct net_device *dev); +extern void netdev_resync_ops(struct net_device *dev); extern int call_netdevice_notifiers(unsigned long val, struct net_device *dev); extern struct net_device *dev_get_by_index(struct net *net, int ifindex); diff --git a/net/8021q/vlan_dev.c b/net/8021q/vlan_dev.c index 4a19acd3a32b..1b34135cf990 100644 --- a/net/8021q/vlan_dev.c +++ b/net/8021q/vlan_dev.c @@ -553,7 +553,7 @@ static int vlan_dev_neigh_setup(struct net_device *dev, struct neigh_parms *pa) int err = 0; if (netif_device_present(real_dev) && ops->ndo_neigh_setup) - err = ops->ndo_neigh_setup(dev, pa); + err = ops->ndo_neigh_setup(real_dev, pa); return err; } @@ -639,6 +639,7 @@ static int vlan_dev_init(struct net_device *dev) dev->hard_header_len = real_dev->hard_header_len + VLAN_HLEN; dev->netdev_ops = &vlan_netdev_ops; } + netdev_resync_ops(dev); if (is_vlan_dev(real_dev)) subclass = 1; diff --git a/net/core/dev.c b/net/core/dev.c index 2dd484ed3dbb..f1129706ce7b 100644 --- a/net/core/dev.c +++ b/net/core/dev.c @@ -4282,6 +4282,39 @@ unsigned long netdev_fix_features(unsigned long features, const char *name) } EXPORT_SYMBOL(netdev_fix_features); +/* Some devices need to (re-)set their netdev_ops inside + * ->init() or similar. If that happens, we have to setup + * the compat pointers again. + */ +void netdev_resync_ops(struct net_device *dev) +{ +#ifdef CONFIG_COMPAT_NET_DEV_OPS + const struct net_device_ops *ops = dev->netdev_ops; + + dev->init = ops->ndo_init; + dev->uninit = ops->ndo_uninit; + dev->open = ops->ndo_open; + dev->change_rx_flags = ops->ndo_change_rx_flags; + dev->set_rx_mode = ops->ndo_set_rx_mode; + dev->set_multicast_list = ops->ndo_set_multicast_list; + dev->set_mac_address = ops->ndo_set_mac_address; + dev->validate_addr = ops->ndo_validate_addr; + dev->do_ioctl = ops->ndo_do_ioctl; + dev->set_config = ops->ndo_set_config; + dev->change_mtu = ops->ndo_change_mtu; + dev->neigh_setup = ops->ndo_neigh_setup; + dev->tx_timeout = ops->ndo_tx_timeout; + dev->get_stats = ops->ndo_get_stats; + dev->vlan_rx_register = ops->ndo_vlan_rx_register; + dev->vlan_rx_add_vid = ops->ndo_vlan_rx_add_vid; + dev->vlan_rx_kill_vid = ops->ndo_vlan_rx_kill_vid; +#ifdef CONFIG_NET_POLL_CONTROLLER + dev->poll_controller = ops->ndo_poll_controller; +#endif +#endif +} +EXPORT_SYMBOL(netdev_resync_ops); + /** * register_netdevice - register a network device * @dev: device to register @@ -4326,28 +4359,7 @@ int register_netdevice(struct net_device *dev) * This is temporary until all network devices are converted. */ if (dev->netdev_ops) { - const struct net_device_ops *ops = dev->netdev_ops; - - dev->init = ops->ndo_init; - dev->uninit = ops->ndo_uninit; - dev->open = ops->ndo_open; - dev->change_rx_flags = ops->ndo_change_rx_flags; - dev->set_rx_mode = ops->ndo_set_rx_mode; - dev->set_multicast_list = ops->ndo_set_multicast_list; - dev->set_mac_address = ops->ndo_set_mac_address; - dev->validate_addr = ops->ndo_validate_addr; - dev->do_ioctl = ops->ndo_do_ioctl; - dev->set_config = ops->ndo_set_config; - dev->change_mtu = ops->ndo_change_mtu; - dev->neigh_setup = ops->ndo_neigh_setup; - dev->tx_timeout = ops->ndo_tx_timeout; - dev->get_stats = ops->ndo_get_stats; - dev->vlan_rx_register = ops->ndo_vlan_rx_register; - dev->vlan_rx_add_vid = ops->ndo_vlan_rx_add_vid; - dev->vlan_rx_kill_vid = ops->ndo_vlan_rx_kill_vid; -#ifdef CONFIG_NET_POLL_CONTROLLER - dev->poll_controller = ops->ndo_poll_controller; -#endif + netdev_resync_ops(dev); } else { char drivername[64]; pr_info("%s (%s): not using net_device_ops yet\n", From 72e2240f181871675d3a979766330c91d48a1673 Mon Sep 17 00:00:00 2001 From: Patrick McHardy Date: Thu, 5 Mar 2009 01:57:44 -0800 Subject: [PATCH 92/92] bonding: Fix device passed into ->ndo_neigh_setup(). Signed-off-by: Patrick McHardy Signed-off-by: David S. Miller --- drivers/net/bonding/bond_main.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/bonding/bond_main.c b/drivers/net/bonding/bond_main.c index 9fb388388fb7..e0578fe8c0db 100644 --- a/drivers/net/bonding/bond_main.c +++ b/drivers/net/bonding/bond_main.c @@ -4113,7 +4113,7 @@ static int bond_neigh_setup(struct net_device *dev, struct neigh_parms *parms) const struct net_device_ops *slave_ops = slave->dev->netdev_ops; if (slave_ops->ndo_neigh_setup) - return slave_ops->ndo_neigh_setup(dev, parms); + return slave_ops->ndo_neigh_setup(slave->dev, parms); } return 0; }