diff --git a/drivers/scsi/elx/efct/efct_driver.c b/drivers/scsi/elx/efct/efct_driver.c index b2b61bc45f12..ae62fc3c9ee3 100644 --- a/drivers/scsi/elx/efct/efct_driver.c +++ b/drivers/scsi/elx/efct/efct_driver.c @@ -261,7 +261,7 @@ efct_firmware_write(struct efct *efct, const u8 *buf, size_t buf_len, dma.size = FW_WRITE_BUFSIZE; dma.virt = dma_alloc_coherent(&efct->pci->dev, - dma.size, &dma.phys, GFP_DMA); + dma.size, &dma.phys, GFP_KERNEL); if (!dma.virt) return -ENOMEM; diff --git a/drivers/scsi/elx/efct/efct_hw.c b/drivers/scsi/elx/efct/efct_hw.c index ba8256b4c782..d4bb37960a3c 100644 --- a/drivers/scsi/elx/efct/efct_hw.c +++ b/drivers/scsi/elx/efct/efct_hw.c @@ -516,7 +516,7 @@ efct_hw_setup_io(struct efct_hw *hw) dma = &hw->xfer_rdy; dma->size = sizeof(struct fcp_txrdy) * hw->config.n_io; dma->virt = dma_alloc_coherent(&efct->pci->dev, - dma->size, &dma->phys, GFP_DMA); + dma->size, &dma->phys, GFP_KERNEL); if (!dma->virt) return -ENOMEM; } @@ -562,7 +562,7 @@ efct_hw_setup_io(struct efct_hw *hw) sizeof(struct sli4_sge); dma->virt = dma_alloc_coherent(&efct->pci->dev, dma->size, &dma->phys, - GFP_DMA); + GFP_KERNEL); if (!dma->virt) { efc_log_err(hw->os, "dma_alloc fail %d\n", i); memset(&io->def_sgl, 0, @@ -618,7 +618,7 @@ efct_hw_init_prereg_io(struct efct_hw *hw) memset(&req, 0, sizeof(struct efc_dma)); req.size = 32 + sgls_per_request * 16; req.virt = dma_alloc_coherent(&efct->pci->dev, req.size, &req.phys, - GFP_DMA); + GFP_KERNEL); if (!req.virt) { kfree(sgls); return -ENOMEM; @@ -1063,7 +1063,7 @@ efct_hw_init(struct efct_hw *hw) dma = &hw->loop_map; dma->size = SLI4_MIN_LOOP_MAP_BYTES; dma->virt = dma_alloc_coherent(&hw->os->pci->dev, dma->size, &dma->phys, - GFP_DMA); + GFP_KERNEL); if (!dma->virt) return -EIO; @@ -1192,7 +1192,7 @@ efct_hw_rx_buffer_alloc(struct efct_hw *hw, u32 rqindex, u32 count, prq->dma.virt = dma_alloc_coherent(&efct->pci->dev, prq->dma.size, &prq->dma.phys, - GFP_DMA); + GFP_KERNEL); if (!prq->dma.virt) { efc_log_err(hw->os, "DMA allocation failed\n"); kfree(rq_buf); diff --git a/drivers/scsi/elx/efct/efct_io.c b/drivers/scsi/elx/efct/efct_io.c index 71e21655916a..c3247b951a76 100644 --- a/drivers/scsi/elx/efct/efct_io.c +++ b/drivers/scsi/elx/efct/efct_io.c @@ -48,7 +48,7 @@ efct_io_pool_create(struct efct *efct, u32 num_sgl) io->rspbuf.size = SCSI_RSP_BUF_LENGTH; io->rspbuf.virt = dma_alloc_coherent(&efct->pci->dev, io->rspbuf.size, - &io->rspbuf.phys, GFP_DMA); + &io->rspbuf.phys, GFP_KERNEL); if (!io->rspbuf.virt) { efc_log_err(efct, "dma_alloc rspbuf failed\n"); efct_io_pool_free(io_pool); diff --git a/drivers/scsi/elx/libefc/efc_cmds.c b/drivers/scsi/elx/libefc/efc_cmds.c index f8665d48904a..da4ac8a4ce12 100644 --- a/drivers/scsi/elx/libefc/efc_cmds.c +++ b/drivers/scsi/elx/libefc/efc_cmds.c @@ -179,7 +179,7 @@ efc_nport_alloc_read_sparm64(struct efc *efc, struct efc_nport *nport) nport->dma.size = EFC_SPARAM_DMA_SZ; nport->dma.virt = dma_alloc_coherent(&efc->pci->dev, nport->dma.size, &nport->dma.phys, - GFP_DMA); + GFP_KERNEL); if (!nport->dma.virt) { efc_log_err(efc, "Failed to allocate DMA memory\n"); efc_nport_free_resources(nport, EFC_EVT_NPORT_ALLOC_FAIL, data); @@ -466,7 +466,7 @@ efc_cmd_domain_alloc(struct efc *efc, struct efc_domain *domain, u32 fcf) domain->dma.size = EFC_SPARAM_DMA_SZ; domain->dma.virt = dma_alloc_coherent(&efc->pci->dev, domain->dma.size, - &domain->dma.phys, GFP_DMA); + &domain->dma.phys, GFP_KERNEL); if (!domain->dma.virt) { efc_log_err(efc, "Failed to allocate DMA memory\n"); return -EIO; diff --git a/drivers/scsi/elx/libefc/efc_els.c b/drivers/scsi/elx/libefc/efc_els.c index 24db0accb256..7bb4f9aad2c8 100644 --- a/drivers/scsi/elx/libefc/efc_els.c +++ b/drivers/scsi/elx/libefc/efc_els.c @@ -71,7 +71,7 @@ efc_els_io_alloc_size(struct efc_node *node, u32 reqlen, u32 rsplen) /* now allocate DMA for request and response */ els->io.req.size = reqlen; els->io.req.virt = dma_alloc_coherent(&efc->pci->dev, els->io.req.size, - &els->io.req.phys, GFP_DMA); + &els->io.req.phys, GFP_KERNEL); if (!els->io.req.virt) { mempool_free(els, efc->els_io_pool); spin_unlock_irqrestore(&node->els_ios_lock, flags); @@ -80,7 +80,7 @@ efc_els_io_alloc_size(struct efc_node *node, u32 reqlen, u32 rsplen) els->io.rsp.size = rsplen; els->io.rsp.virt = dma_alloc_coherent(&efc->pci->dev, els->io.rsp.size, - &els->io.rsp.phys, GFP_DMA); + &els->io.rsp.phys, GFP_KERNEL); if (!els->io.rsp.virt) { dma_free_coherent(&efc->pci->dev, els->io.req.size, els->io.req.virt, els->io.req.phys); diff --git a/drivers/scsi/elx/libefc_sli/sli4.c b/drivers/scsi/elx/libefc_sli/sli4.c index 907d67aeac23..3ea57bd6fb0a 100644 --- a/drivers/scsi/elx/libefc_sli/sli4.c +++ b/drivers/scsi/elx/libefc_sli/sli4.c @@ -445,7 +445,7 @@ sli_cmd_rq_create_v2(struct sli4 *sli4, u32 num_rqs, dma->size = payload_size; dma->virt = dma_alloc_coherent(&sli4->pci->dev, dma->size, - &dma->phys, GFP_DMA); + &dma->phys, GFP_KERNEL); if (!dma->virt) return -EIO; @@ -508,7 +508,7 @@ __sli_queue_init(struct sli4 *sli4, struct sli4_queue *q, u32 qtype, q->dma.size = size * n_entries; q->dma.virt = dma_alloc_coherent(&sli4->pci->dev, q->dma.size, - &q->dma.phys, GFP_DMA); + &q->dma.phys, GFP_KERNEL); if (!q->dma.virt) { memset(&q->dma, 0, sizeof(struct efc_dma)); efc_log_err(sli4, "%s allocation failed\n", SLI4_QNAME[qtype]); @@ -849,7 +849,7 @@ static int sli_cmd_cq_set_create(struct sli4 *sli4, dma->size = payload_size; dma->virt = dma_alloc_coherent(&sli4->pci->dev, dma->size, - &dma->phys, GFP_DMA); + &dma->phys, GFP_KERNEL); if (!dma->virt) return -EIO; @@ -4413,7 +4413,7 @@ sli_get_ctrl_attributes(struct sli4 *sli4) psize = sizeof(struct sli4_rsp_cmn_get_cntl_addl_attributes); data.size = psize; data.virt = dma_alloc_coherent(&sli4->pci->dev, data.size, - &data.phys, GFP_DMA); + &data.phys, GFP_KERNEL); if (!data.virt) { memset(&data, 0, sizeof(struct efc_dma)); efc_log_err(sli4, "Failed to allocate memory for GET_CNTL_ADDL_ATTR\n"); @@ -4653,7 +4653,7 @@ sli_setup(struct sli4 *sli4, void *os, struct pci_dev *pdev, */ sli4->bmbx.size = SLI4_BMBX_SIZE + sizeof(struct sli4_mcqe); sli4->bmbx.virt = dma_alloc_coherent(&pdev->dev, sli4->bmbx.size, - &sli4->bmbx.phys, GFP_DMA); + &sli4->bmbx.phys, GFP_KERNEL); if (!sli4->bmbx.virt) { memset(&sli4->bmbx, 0, sizeof(struct efc_dma)); efc_log_err(sli4, "bootstrap mailbox allocation failed\n"); @@ -4674,7 +4674,7 @@ sli_setup(struct sli4 *sli4, void *os, struct pci_dev *pdev, sli4->vpd_data.virt = dma_alloc_coherent(&pdev->dev, sli4->vpd_data.size, &sli4->vpd_data.phys, - GFP_DMA); + GFP_KERNEL); if (!sli4->vpd_data.virt) { memset(&sli4->vpd_data, 0, sizeof(struct efc_dma)); /* Note that failure isn't fatal in this specific case */ @@ -5070,7 +5070,7 @@ sli_cmd_post_hdr_templates(struct sli4 *sli4, void *buf, struct efc_dma *dma, payload_dma->size = payload_size; payload_dma->virt = dma_alloc_coherent(&sli4->pci->dev, payload_dma->size, - &payload_dma->phys, GFP_DMA); + &payload_dma->phys, GFP_KERNEL); if (!payload_dma->virt) { memset(payload_dma, 0, sizeof(struct efc_dma)); efc_log_err(sli4, "mbox payload memory allocation fail\n");