Start using the helpers that align buffer object user-space addresses and
buffer object vram addresses to huge page boundaries.
This is to improve the chances of allowing huge page-table entries.
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: "Jérôme Glisse" <jglisse@redhat.com>
Cc: "Christian König" <christian.koenig@amd.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Thomas Hellstrom (VMware) <thomas_os@shipmail.org>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Acked-by: Christian König <christian.koenig@amd.com>
Unaligned virtual addresses makes it unlikely that huge page-table entries
can be used.
So align virtual buffer object address huge page boundaries to the
underlying physical address huge page boundaries taking buffer object
sizes into account to determine when it might be possible to use huge
page-table entries.
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Michal Hocko <mhocko@suse.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Ralph Campbell <rcampbell@nvidia.com>
Cc: "Jérôme Glisse" <jglisse@redhat.com>
Cc: "Christian König" <christian.koenig@amd.com>
Cc: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: Thomas Hellstrom (VMware) <thomas_os@shipmail.org>
Reviewed-by: Roland Scheidegger <sroland@vmware.com>
Acked-by: Christian König <christian.koenig@amd.com>
This catches the majority of drivers (unfortunately not if we take
users into account, because all the big drivers have at least a
lastclose hook).
With the prep patches out of the way all drm state is fully protected
and either prevents or can deal with the races from dropping the BKL
around open/close. The only thing left to audit are the various driver
hooks - by keeping the BKL around if any of them are set we have a
very simple cop-out!
Note that one of the biggest prep pieces to get here was making
dev->open_count atomic, which was done in
commit 7e13ad8964
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date: Fri Jan 24 13:01:07 2020 +0000
drm: Avoid drm_global_mutex for simple inc/dec of dev->open_count
v2:
- Rebase and fix locking in drm_open() (Chris)
- Indentation fix in drm_release
- Typo fix in the commit message (Sam)
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Sam Ravnborg <sam@ravnborg.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Tested-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200204150146.2006481-6-daniel.vetter@ffwll.ch
We want to only take the BKL on crap drivers, but to know whether
we have a crap driver we first need to look it up. Split this shuffle
out from the main BKL-disabling patch, for more clarity. Historical
aside: When the kernel-wide BKL was removed, it was replaced by
drm_global_mutex within the scope of the drm subsystem hence why these
two things are (almost) interchangeable as concepts here.
Since the minors are refcounted drm_minor_acquire is purely internal
and this does not have a driver visible effect.
v2: Push the locking even further into drm_open(), suggested by Chris.
This gives us more symmetry with drm_release(), and maybe a futuer
avenue where we make drm_global_mutex locking (partially) opt-in like
with drm_release_noglobal().
v3:
- Actually push this stuff correctly, don't unlock twice (Chris)
- Fix typo on commit message, plus explain why BKL = drm_global_mutex
(Sam)
Cc: Sam Ravnborg <sam@ravnborg.org>
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Tested-by: Thomas Zimmermann <tzimmermann@suse.de>
Acked-by: Thomas Zimmermann <tzimmermann@suse.de>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200204150146.2006481-5-daniel.vetter@ffwll.ch
Since drm_global_mutex is a true global mutex across devices, we don't
want to acquire it unless absolutely necessary. For maintaining the
device local open_count, we can use atomic operations on the counter
itself, except when making the transition to/from 0. Here, we tackle the
easy portion of delaying acquiring the drm_global_mutex for the final
release by using atomic_dec_and_mutex_lock(), leaving the global
serialisation across the device opens.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Thomas Hellström (VMware) <thomas_os@shipmail.org>
Reviewed-by: Thomas Hellström <thellstrom@vmware.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200124130107.125404-1-chris@chris-wilson.co.uk
The file is not part of the global drm resource and can be released
prior to take the global mutex to drop the open_count (and potentially
close) the drm device. As the global mutex is indeed global, not only
within the device but across devices, a slow file release mechanism can
bottleneck the entire system.
However, inside drm_close_helper() there are a number of dev->driver
callbacks that take the drm_device as the first parameter... Worryingly
some of those callbacks may be (implicitly) depending on the global
mutex.
v2: Drop the debug message for the open-count, it's included with the
drm_file_free() debug message -- and for good measure make that up as
reading outside of the mutex.
v3: Separate the calling of the filp cleanup outside of
drm_global_mutex into a new drm_release_noglobal() hook, so that we can
phase the transition. drm/savage relies on the global mutex, and there
may be more, so be cautious.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Thomas Hellström (VMware) <thomas_os@shipmail.org>
Reviewed-by: Thomas Hellström (VMware) <thomas_os@shipmail.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20200124125627.125042-1-chris@chris-wilson.co.uk
Sometimes we need to create a struct file to wrap a drm_device, as it
the user were to have opened /dev/dri/card0 but to do so anonymously
(i.e. for internal use). Provide a utility method to create a struct
file with the drm_device->driver.fops, that wrap the drm_device.
v2: Restrict usage to selftests
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20191107180601.30815-2-chris@chris-wilson.co.uk
The use of the drmP.h header file is deprecated.
Remove use from all files in drm/*
so people do not look there and follow a bad example.
Build tested allyesconfig,allmodconfig on x86, arm etc.
Including alpha that is as always more challenging than
the rest.
Signed-off-by: Sam Ravnborg <sam@ravnborg.org>
Acked-by: Daniel Vetter <daniel@ffwll.ch>
Cc: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Cc: Maxime Ripard <maxime.ripard@bootlin.com>
Cc: Sean Paul <sean@poorly.run>
Cc: David Airlie <airlied@linux.ie>
Link: https://patchwork.freedesktop.org/patch/msgid/20190526173535.32701-8-sam@ravnborg.org
Move the open helper around to avoid the forward decl, and give
drm_setup a drm_legacy_ prefix since it's all legacy stuff in there.
v2: Move drm_legacy_setup into drm_legacy_misc.c (Chris). The
counterpart in the form of drm_legacy_dev_reinit is there already too,
plus it fits perfectly into Dave's work of making DRIVER_LEGACY code
compile-time optional.
Cc: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20190502135603.20413-1-daniel.vetter@ffwll.ch
This removes these unless legacy is enabled.
The lock count init is unneeded anyways since it's kzalloc.
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
This moves the legacy dev reinit into the legacy misc file.
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Dave Airlie <airlied@redhat.com>
- Block fb changes for async atomic updates to prevent a use after free.
- Fix ID mismatch error on load in bochs.
- Fix memory leak when drm_setup fails.
- Fixes around handling of DRM_AUTH.
-
-----BEGIN PGP SIGNATURE-----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=Rf1K
-----END PGP SIGNATURE-----
Merge tag 'drm-misc-fixes-2019-02-22' of git://anongit.freedesktop.org/drm/drm-misc into drm-next
drm-misc-fixes for v5.0:
- Block fb changes for async atomic updates to prevent a use after free.
- Fix ID mismatch error on load in bochs.
- Fix memory leak when drm_setup fails.
- Fixes around handling of DRM_AUTH.
Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/42113611-e2cd-6bdd-7de5-4f8ab5a0cbe6@linux.intel.com
After an event is sent, we try to copy it into the user buffer of the
first waiter in drm_read() and if the user buffer doesn't have enough
room we put it back onto the list. However, we didn't wake up any
subsequent waiter, so that event may sit on the list until either a new
vblank event is sent or a new waiter appears. Rare, but in the worst
case may lead to a stuck process.
Testcase: igt/drm_read/short-buffer-wakeup
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170804082328.17173-1-chris@chris-wilson.co.uk
If userspace has open fd(s) when drm_dev_unplug() is run, it will result
in drm_dev_unregister() being called twice. First in drm_dev_unplug() and
then later in drm_release() through the call to drm_put_dev().
Since userspace already holds a ref on drm_device through the drm_minor,
it's not necessary to add extra ref counting based on no open file
handles. Instead just drm_dev_put() unconditionally in drm_dev_unplug().
We now have this:
- Userpace holds a ref on drm_device as long as there's open fd(s)
- The driver holds a ref on drm_device as long as it's bound to the
struct device
When both sides are done with drm_device, it is released.
Signed-off-by: Noralf Trønnes <noralf@tronnes.org>
Reviewed-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Sean Paul <sean@poorly.run>
Link: https://patchwork.freedesktop.org/patch/msgid/20190208140103.28919-2-noralf@tronnes.org
Nobody has actually used the type (VERIFY_READ vs VERIFY_WRITE) argument
of the user address range verification function since we got rid of the
old racy i386-only code to walk page tables by hand.
It existed because the original 80386 would not honor the write protect
bit when in kernel mode, so you had to do COW by hand before doing any
user access. But we haven't supported that in a long time, and these
days the 'type' argument is a purely historical artifact.
A discussion about extending 'user_access_begin()' to do the range
checking resulted this patch, because there is no way we're going to
move the old VERIFY_xyz interface to that model. And it's best done at
the end of the merge window when I've done most of my merges, so let's
just get this done once and for all.
This patch was mostly done with a sed-script, with manual fix-ups for
the cases that weren't of the trivial 'access_ok(VERIFY_xyz' form.
There were a couple of notable cases:
- csky still had the old "verify_area()" name as an alias.
- the iter_iov code had magical hardcoded knowledge of the actual
values of VERIFY_{READ,WRITE} (not that they mattered, since nothing
really used it)
- microblaze used the type argument for a debug printout
but other than those oddities this should be a total no-op patch.
I tried to fix up all architectures, did fairly extensive grepping for
access_ok() uses, and the changes are trivial, but I may have missed
something. Any missed conversion should be trivially fixable, though.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This the beginning of an API for in-kernel clients.
First out is a way to get a framebuffer backed by a dumb buffer.
Only GEM drivers are supported.
The original idea of using an exported dma-buf was dropped because it
also creates an anonomous file descriptor which doesn't work when the
buffer is created from a kernel thread. The easy way out is to use
drm_driver.gem_prime_vmap to get the virtual address, which requires a
GEM object. This excludes the vmwgfx driver which is the only non-GEM
driver apart from the legacy ones. A solution for vmwgfx will have to be
worked out later if it wants to support the client API which it probably
will when we have a bootsplash client.
Suggested-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Signed-off-by: Noralf Trønnes <noralf@tronnes.org>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20180703160354.59955-2-noralf@tronnes.org
Rather than doing drm_file allocation/destruction right in the fops, lets
provide separate helpers. This decouples drm_file management from the
still-mandatory drm-fops. It prepares for use of drm_file without the
fops, both by possible separate fops implementations and APIs (not that I
am aware of any such plans), and more importantly from in-kernel use where
no real file is available.
Signed-off-by: David Herrmann <dh.herrmann@gmail.com>
Signed-off-by: Noralf Trønnes <noralf@tronnes.org>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Link: https://patchwork.freedesktop.org/patch/msgid/20180618141739.48151-2-noralf@tronnes.org
Since we have the ttm and gem vma managers using a subset
of the file address space for objects, and these start at
0x100000000 they will overflow the new mmap checks.
I've checked all the mmap routines I could see for any
bad behaviour but overall most people use GEM/TTM VMA
managers even the legacy drivers have a hashtable.
Reported-and-Tested-by: Arthur Marsh (amarsh04 on #radeon)
Fixes: be83bbf806 (mmap: introduce sane default mmap limits)
Signed-off-by: Dave Airlie <airlied@redhat.com>
This is the mindless scripted replacement of kernel use of POLL*
variables as described by Al, done by this script:
for V in IN OUT PRI ERR RDNORM RDBAND WRNORM WRBAND HUP RDHUP NVAL MSG; do
L=`git grep -l -w POLL$V | grep -v '^t' | grep -v /um/ | grep -v '^sa' | grep -v '/poll.h$'|grep -v '^D'`
for f in $L; do sed -i "-es/^\([^\"]*\)\(\<POLL$V\>\)/\\1E\\2/" $f; done
done
with de-mangling cleanups yet to come.
NOTE! On almost all architectures, the EPOLL* constants have the same
values as the POLL* constants do. But they keyword here is "almost".
For various bad reasons they aren't the same, and epoll() doesn't
actually work quite correctly in some cases due to this on Sparc et al.
The next patch from Al will sort out the final differences, and we
should be all done.
Scripted-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
While at it, also ocd and give them a consistent drm_dev_ prefix, like
the other device instance functionality. Plus move the functions into
the right places.
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20170802115604.12734-3-daniel.vetter@ffwll.ch
I spotted a markup issue, plus adding the descriptions in drm_driver.
Plus a few more links while at it.
I'm still mildly unhappy with the split between fops and ioctls, but I
still think having the ioctls in the uapi chapter makes more sense. Oh
well ...
v2: Rebase.
v3: Move misplace hunk to the right patch.
Cc: Stefan Agner <stefan@agner.ch>
Acked-by: Thierry Reding <treding@nvidia.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170531092045.3950-1-daniel.vetter@ffwll.ch
Sync objects are new toplevel drm object, that contain a
pointer to a fence. This fence can be updated via command
submission ioctls via drivers.
There is also a generic wait obj API modelled on the vulkan
wait API (with code modelled on some amdgpu code).
These objects can be converted to an opaque fd that can be
passes between processes.
v2: rename reference/unreference to put/get (Chris)
fix leaked reference (David Zhou)
drop mutex in favour of cmpxchg (Chris)
v3: cleanups from danvet, rebase on drm_fops rename
check fd_flags is 0 in ioctls.
v4: export find/free, change replace fence to take a
syncobj. In order to support lookup first, replace
later semantics which seem in the end to be cleaner.
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Dave Airlie <airlied@redhat.com>
With all drivers converted there's only legacy dri1 drivers using it.
Not going to touch those, instead just hide it like we've done with
other dri1 driver hooks like firstopen.
In all this I didn't find any real reason why we'd needed 2 hooks, and
having symmetry between open and close just appeases my OCD better.
Yeah, someone else could do an s/postclose/close/, but that's for
someone who understands cocci. And maybe after this series is reviewed
and landed, to avoid patch-regen churn.
v2: s/last/post/close in the kernel-doc (Sean).
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Liviu Dudau <Liviu.Dudau@arm.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170508082633.4214-4-daniel.vetter@ffwll.ch
Sadly there's only 1 driver which can use it, everyone else is special
for some reason:
- gma500 has a horrible runtime PM ioctl wrapper that probably doesn't
really work but meh.
- i915 needs special compat_ioctl handler because regrets.
- arcgpu needs to fixup the pgprot because (no idea why it can't do
that in the fault handler like everyone else).
- tegra does even worse stuff with pgprot
- udl does something with vm_flags too ...
- cma helpers, etnaviv, mtk, msm, rockchip, omap all implement some
variation on prefaulting.
- exynos is exynos, I got lost in the midlayers.
- vc4 has to reinvent half of cma helpers because those are too much
midlayer, plus vm_flags dances.
- vgem also seems unhappy with the default vm_flags.
So pretty sad divergence and I'm sure we could do better, but not
really an idea. Oh well, maybe this macro here helps to encourage more
consistency at least going forward.
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Liviu Dudau <Liviu.Dudau@arm.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170308141257.12119-25-daniel.vetter@ffwll.ch
Less code ftw.
This converts all drivers except the tinydrm helper module. That one
needs more work, since it gets the THIS_MODULE reference from
tinydrm.ko instead of the actual driver module like it should.
Probably needs a similar trick like I used here with generating the
entire struct with a macro.
Cc: Noralf Trønnes <noralf@tronnes.org>
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Liviu Dudau <Liviu.Dudau@arm.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170308141257.12119-24-daniel.vetter@ffwll.ch
Well, mostly drm_file.h, and clean up all related things:
- I didnt' figure out the difference between preclose and postclose.
The existing explanation in drm-internals.rst didn't convince me,
since it's also really outdated - we clean up pending DRM events in
the core nowadays. I put a FIXME in for the future.
- Another FIXME is to have a macro for default fops.
- Lots of links all around, main areas are to tie the overview in
drm_file.c more into the callbacks in struct drm_device, and the
other is to link render/primary node code to the right sections in
drm-uapi.rst.
- Also moved the open/close stuff to drm_drv.h from drm-internals.rst,
seems like the better place for that information. Since that section
was rather outdated this amounted to full-on rewrite.
A big missing piece here is some overview graph, but I think better to
wait with that one until drm_device and drm_driver are also fully
documented.
v2: Nits from Sean.
Reviewed-by: Sean Paul <seanpaul@chromium.org>
Reviewed-by: Liviu Dudau <Liviu.Dudau@arm.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170308141257.12119-12-daniel.vetter@ffwll.ch
I'm torn on whether drm_minor really should be here or somewhere else.
Maybe with more clarity after untangling drmP.h more this is easier to
decide, for now I've put a FIXME comment right next to it. Right now
we need struct drm_minor for the inline drm_file type helpers, and so
it does kinda make sense to have them here.
Next patch will kerneldoc-ify the entire pile.
Reviewed-by: Gustavo Padovan <gustavo.padovan@collabora.com>
Signed-off-by: Daniel Vetter <daniel.vetter@intel.com>
Link: http://patchwork.freedesktop.org/patch/msgid/20170308141257.12119-10-daniel.vetter@ffwll.ch
Thomas Hellstrom (VMware)
Daniel Vetter
Chris Wilson
Dave Airlie
Emil Velikov
Sam Ravnborg
Sean Paul
Noralf Trønnes
Matt Roper
Linus Torvalds
David Herrmann
Al Viro