I/OAT: code cleanup from checkpatch output
authorShannon Nelson <shannon.nelson@intel.com>
Tue, 16 Oct 2007 08:27:39 +0000 (01:27 -0700)
committerLinus Torvalds <torvalds@woody.linux-foundation.org>
Tue, 16 Oct 2007 16:43:09 +0000 (09:43 -0700)
Take care of a bunch of little code nits in ioatdma files

Signed-off-by: Shannon Nelson <shannon.nelson@intel.com>
Acked-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
drivers/dma/ioat_dma.c

index f0be71c..2db05f6 100644 (file)
@@ -1,10 +1,10 @@
 /*
- * Copyright(c) 2004 - 2006 Intel Corporation. All rights reserved.
+ * Intel I/OAT DMA Linux driver
+ * Copyright(c) 2004 - 2007 Intel Corporation.
  *
  * This program is free software; you can redistribute it and/or modify it
- * under the terms of the GNU General Public License as published by the Free
- * Software Foundation; either version 2 of the License, or (at your option)
- * any later version.
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
  *
  * This program is distributed in the hope that it will be useful, but WITHOUT
  * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
  * more details.
  *
  * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc., 59
- * Temple Place - Suite 330, Boston, MA  02111-1307, USA.
+ * this program; if not, write to the Free Software Foundation, Inc.,
+ * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
+ *
+ * The full GNU General Public License is included in this distribution in
+ * the file called "COPYING".
  *
- * The full GNU General Public License is included in this distribution in the
- * file called COPYING.
  */
 
 /*
 #include "ioatdma_registers.h"
 #include "ioatdma_hw.h"
 
+#define INITIAL_IOAT_DESC_COUNT 128
+
 #define to_ioat_chan(chan) container_of(chan, struct ioat_dma_chan, common)
 #define to_ioat_device(dev) container_of(dev, struct ioat_device, common)
 #define to_ioat_desc(lh) container_of(lh, struct ioat_desc_sw, node)
 #define tx_to_ioat_desc(tx) container_of(tx, struct ioat_desc_sw, async_tx)
 
 /* internal functions */
-static int __devinit ioat_probe(struct pci_dev *pdev, const struct pci_device_id *ent);
+static void ioat_dma_start_null_desc(struct ioat_dma_chan *ioat_chan);
+static void ioat_dma_memcpy_cleanup(struct ioat_dma_chan *ioat_chan);
+static int __devinit ioat_probe(struct pci_dev *pdev,
+                               const struct pci_device_id *ent);
 static void ioat_shutdown(struct pci_dev *pdev);
 static void __devexit ioat_remove(struct pci_dev *pdev);
 
-static int enumerate_dma_channels(struct ioat_device *device)
+static int ioat_dma_enumerate_channels(struct ioat_device *device)
 {
        u8 xfercap_scale;
        u32 xfercap;
@@ -73,13 +79,14 @@ static int enumerate_dma_channels(struct ioat_device *device)
                /* This should be made common somewhere in dmaengine.c */
                ioat_chan->common.device = &device->common;
                list_add_tail(&ioat_chan->common.device_node,
-                             &device->common.channels);
+                             &device->common.channels);
        }
        return device->common.chancnt;
 }
 
-static void
-ioat_set_src(dma_addr_t addr, struct dma_async_tx_descriptor *tx, int index)
+static void ioat_set_src(dma_addr_t addr,
+                        struct dma_async_tx_descriptor *tx,
+                        int index)
 {
        struct ioat_desc_sw *iter, *desc = tx_to_ioat_desc(tx);
        struct ioat_dma_chan *ioat_chan = to_ioat_chan(tx->chan);
@@ -93,8 +100,9 @@ ioat_set_src(dma_addr_t addr, struct dma_async_tx_descriptor *tx, int index)
 
 }
 
-static void
-ioat_set_dest(dma_addr_t addr, struct dma_async_tx_descriptor *tx, int index)
+static void ioat_set_dest(dma_addr_t addr,
+                         struct dma_async_tx_descriptor *tx,
+                         int index)
 {
        struct ioat_desc_sw *iter, *desc = tx_to_ioat_desc(tx);
        struct ioat_dma_chan *ioat_chan = to_ioat_chan(tx->chan);
@@ -107,8 +115,7 @@ ioat_set_dest(dma_addr_t addr, struct dma_async_tx_descriptor *tx, int index)
        }
 }
 
-static dma_cookie_t
-ioat_tx_submit(struct dma_async_tx_descriptor *tx)
+static dma_cookie_t ioat_tx_submit(struct dma_async_tx_descriptor *tx)
 {
        struct ioat_dma_chan *ioat_chan = to_ioat_chan(tx->chan);
        struct ioat_desc_sw *desc = tx_to_ioat_desc(tx);
@@ -146,8 +153,8 @@ ioat_tx_submit(struct dma_async_tx_descriptor *tx)
 }
 
 static struct ioat_desc_sw *ioat_dma_alloc_descriptor(
-       struct ioat_dma_chan *ioat_chan,
-       gfp_t flags)
+                                       struct ioat_dma_chan *ioat_chan,
+                                       gfp_t flags)
 {
        struct ioat_dma_descriptor *desc;
        struct ioat_desc_sw *desc_sw;
@@ -177,10 +184,6 @@ static struct ioat_desc_sw *ioat_dma_alloc_descriptor(
        return desc_sw;
 }
 
-#define INITIAL_IOAT_DESC_COUNT 128
-
-static void ioat_start_null_desc(struct ioat_dma_chan *ioat_chan);
-
 /* returns the actual number of allocated descriptors */
 static int ioat_dma_alloc_chan_resources(struct dma_chan *chan)
 {
@@ -195,15 +198,16 @@ static int ioat_dma_alloc_chan_resources(struct dma_chan *chan)
        if (!list_empty(&ioat_chan->free_desc))
                return INITIAL_IOAT_DESC_COUNT;
 
-        /* Setup register to interrupt and write completion status on error */
+       /* Setup register to interrupt and write completion status on error */
        chanctrl = IOAT_CHANCTRL_ERR_INT_EN |
                IOAT_CHANCTRL_ANY_ERR_ABORT_EN |
                IOAT_CHANCTRL_ERR_COMPLETION_EN;
-        writew(chanctrl, ioat_chan->reg_base + IOAT_CHANCTRL_OFFSET);
+       writew(chanctrl, ioat_chan->reg_base + IOAT_CHANCTRL_OFFSET);
 
        chanerr = readl(ioat_chan->reg_base + IOAT_CHANERR_OFFSET);
        if (chanerr) {
-               printk("IOAT: CHANERR = %x, clearing\n", chanerr);
+               dev_err(&ioat_chan->device->pdev->dev,
+                       "ioatdma: CHANERR = %x, clearing\n", chanerr);
                writel(chanerr, ioat_chan->reg_base + IOAT_CHANERR_OFFSET);
        }
 
@@ -211,7 +215,8 @@ static int ioat_dma_alloc_chan_resources(struct dma_chan *chan)
        for (i = 0; i < INITIAL_IOAT_DESC_COUNT; i++) {
                desc = ioat_dma_alloc_descriptor(ioat_chan, GFP_KERNEL);
                if (!desc) {
-                       printk(KERN_ERR "IOAT: Only %d initial descriptors\n", i);
+                       dev_err(&ioat_chan->device->pdev->dev,
+                               "ioatdma: Only %d initial descriptors\n", i);
                        break;
                }
                list_add_tail(&desc->node, &tmp_list);
@@ -224,8 +229,8 @@ static int ioat_dma_alloc_chan_resources(struct dma_chan *chan)
        /* doing 2 32bit writes to mmio since 1 64b write doesn't work */
        ioat_chan->completion_virt =
                pci_pool_alloc(ioat_chan->device->completion_pool,
-                              GFP_KERNEL,
-                              &ioat_chan->completion_addr);
+                              GFP_KERNEL,
+                              &ioat_chan->completion_addr);
        memset(ioat_chan->completion_virt, 0,
               sizeof(*ioat_chan->completion_virt));
        writel(((u64) ioat_chan->completion_addr) & 0x00000000FFFFFFFF,
@@ -233,12 +238,10 @@ static int ioat_dma_alloc_chan_resources(struct dma_chan *chan)
        writel(((u64) ioat_chan->completion_addr) >> 32,
               ioat_chan->reg_base + IOAT_CHANCMP_OFFSET_HIGH);
 
-       ioat_start_null_desc(ioat_chan);
+       ioat_dma_start_null_desc(ioat_chan);
        return i;
 }
 
-static void ioat_dma_memcpy_cleanup(struct ioat_dma_chan *ioat_chan);
-
 static void ioat_dma_free_chan_resources(struct dma_chan *chan)
 {
        struct ioat_dma_chan *ioat_chan = to_ioat_chan(chan);
@@ -268,19 +271,22 @@ static void ioat_dma_free_chan_resources(struct dma_chan *chan)
        spin_unlock_bh(&ioat_chan->desc_lock);
 
        pci_pool_free(ioat_device->completion_pool,
-                     ioat_chan->completion_virt,
-                     ioat_chan->completion_addr);
+                     ioat_chan->completion_virt,
+                     ioat_chan->completion_addr);
 
        /* one is ok since we left it on there on purpose */
        if (in_use_descs > 1)
-               printk(KERN_ERR "IOAT: Freeing %d in use descriptors!\n",
+               dev_err(&ioat_chan->device->pdev->dev,
+                       "ioatdma: Freeing %d in use descriptors!\n",
                        in_use_descs - 1);
 
        ioat_chan->last_completion = ioat_chan->completion_addr = 0;
 }
 
-static struct dma_async_tx_descriptor *
-ioat_dma_prep_memcpy(struct dma_chan *chan, size_t len, int int_en)
+static struct dma_async_tx_descriptor *ioat_dma_prep_memcpy(
+                                               struct dma_chan *chan,
+                                               size_t len,
+                                               int int_en)
 {
        struct ioat_dma_chan *ioat_chan = to_ioat_chan(chan);
        struct ioat_desc_sw *first, *prev, *new;
@@ -343,12 +349,11 @@ ioat_dma_prep_memcpy(struct dma_chan *chan, size_t len, int int_en)
        return new ? &new->async_tx : NULL;
 }
 
-
 /**
- * ioat_dma_memcpy_issue_pending - push potentially unrecognized appended descriptors to hw
+ * ioat_dma_memcpy_issue_pending - push potentially unrecognized appended
+ *                                 descriptors to hw
  * @chan: DMA channel handle
  */
-
 static void ioat_dma_memcpy_issue_pending(struct dma_chan *chan)
 {
        struct ioat_dma_chan *ioat_chan = to_ioat_chan(chan);
@@ -360,15 +365,15 @@ static void ioat_dma_memcpy_issue_pending(struct dma_chan *chan)
        }
 }
 
-static void ioat_dma_memcpy_cleanup(struct ioat_dma_chan *chan)
+static void ioat_dma_memcpy_cleanup(struct ioat_dma_chan *ioat_chan)
 {
        unsigned long phys_complete;
        struct ioat_desc_sw *desc, *_desc;
        dma_cookie_t cookie = 0;
 
-       prefetch(chan->completion_virt);
+       prefetch(ioat_chan->completion_virt);
 
-       if (!spin_trylock(&chan->cleanup_lock))
+       if (!spin_trylock(&ioat_chan->cleanup_lock))
                return;
 
        /* The completion writeback can happen at any time,
@@ -378,26 +383,27 @@ static void ioat_dma_memcpy_cleanup(struct ioat_dma_chan *chan)
 
 #if (BITS_PER_LONG == 64)
        phys_complete =
-       chan->completion_virt->full & IOAT_CHANSTS_COMPLETED_DESCRIPTOR_ADDR;
+       ioat_chan->completion_virt->full & IOAT_CHANSTS_COMPLETED_DESCRIPTOR_ADDR;
 #else
-       phys_complete = chan->completion_virt->low & IOAT_LOW_COMPLETION_MASK;
+       phys_complete = ioat_chan->completion_virt->low & IOAT_LOW_COMPLETION_MASK;
 #endif
 
-       if ((chan->completion_virt->full & IOAT_CHANSTS_DMA_TRANSFER_STATUS) ==
-               IOAT_CHANSTS_DMA_TRANSFER_STATUS_HALTED) {
-               printk("IOAT: Channel halted, chanerr = %x\n",
-                       readl(chan->reg_base + IOAT_CHANERR_OFFSET));
+       if ((ioat_chan->completion_virt->full & IOAT_CHANSTS_DMA_TRANSFER_STATUS) ==
+                               IOAT_CHANSTS_DMA_TRANSFER_STATUS_HALTED) {
+               dev_err(&ioat_chan->device->pdev->dev,
+                       "ioatdma: Channel halted, chanerr = %x\n",
+                       readl(ioat_chan->reg_base + IOAT_CHANERR_OFFSET));
 
                /* TODO do something to salvage the situation */
        }
 
-       if (phys_complete == chan->last_completion) {
-               spin_unlock(&chan->cleanup_lock);
+       if (phys_complete == ioat_chan->last_completion) {
+               spin_unlock(&ioat_chan->cleanup_lock);
                return;
        }
 
-       spin_lock_bh(&chan->desc_lock);
-       list_for_each_entry_safe(desc, _desc, &chan->used_desc, node) {
+       spin_lock_bh(&ioat_chan->desc_lock);
+       list_for_each_entry_safe(desc, _desc, &ioat_chan->used_desc, node) {
 
                /*
                 * Incoming DMA requests may use multiple descriptors, due to
@@ -407,31 +413,36 @@ static void ioat_dma_memcpy_cleanup(struct ioat_dma_chan *chan)
                if (desc->async_tx.cookie) {
                        cookie = desc->async_tx.cookie;
 
-                       /* yes we are unmapping both _page and _single alloc'd
-                          regions with unmap_page. Is this *really* that bad?
-                       */
-                       pci_unmap_page(chan->device->pdev,
+                       /*
+                        * yes we are unmapping both _page and _single alloc'd
+                        * regions with unmap_page. Is this *really* that bad?
+                        */
+                       pci_unmap_page(ioat_chan->device->pdev,
                                        pci_unmap_addr(desc, dst),
                                        pci_unmap_len(desc, len),
                                        PCI_DMA_FROMDEVICE);
-                       pci_unmap_page(chan->device->pdev,
+                       pci_unmap_page(ioat_chan->device->pdev,
                                        pci_unmap_addr(desc, src),
                                        pci_unmap_len(desc, len),
                                        PCI_DMA_TODEVICE);
                }
 
                if (desc->async_tx.phys != phys_complete) {
-                       /* a completed entry, but not the last, so cleanup
+                       /*
+                        * a completed entry, but not the last, so cleanup
                         * if the client is done with the descriptor
                         */
                        if (desc->async_tx.ack) {
                                list_del(&desc->node);
-                               list_add_tail(&desc->node, &chan->free_desc);
+                               list_add_tail(&desc->node,
+                                             &ioat_chan->free_desc);
                        } else
                                desc->async_tx.cookie = 0;
                } else {
-                       /* last used desc. Do not remove, so we can append from
-                          it, but don't look at it next time, either */
+                       /*
+                        * last used desc. Do not remove, so we can append from
+                        * it, but don't look at it next time, either
+                        */
                        desc->async_tx.cookie = 0;
 
                        /* TODO check status bits? */
@@ -439,13 +450,13 @@ static void ioat_dma_memcpy_cleanup(struct ioat_dma_chan *chan)
                }
        }
 
-       spin_unlock_bh(&chan->desc_lock);
+       spin_unlock_bh(&ioat_chan->desc_lock);
 
-       chan->last_completion = phys_complete;
+       ioat_chan->last_completion = phys_complete;
        if (cookie != 0)
-               chan->completed_cookie = cookie;
+               ioat_chan->completed_cookie = cookie;
 
-       spin_unlock(&chan->cleanup_lock);
+       spin_unlock(&ioat_chan->cleanup_lock);
 }
 
 static void ioat_dma_dependency_added(struct dma_chan *chan)
@@ -466,11 +477,10 @@ static void ioat_dma_dependency_added(struct dma_chan *chan)
  * @done: if not %NULL, updated with last completed transaction
  * @used: if not %NULL, updated with last used transaction
  */
-
 static enum dma_status ioat_dma_is_complete(struct dma_chan *chan,
-                                            dma_cookie_t cookie,
-                                            dma_cookie_t *done,
-                                            dma_cookie_t *used)
+                                           dma_cookie_t cookie,
+                                           dma_cookie_t *done,
+                                           dma_cookie_t *used)
 {
        struct ioat_dma_chan *ioat_chan = to_ioat_chan(chan);
        dma_cookie_t last_used;
@@ -481,7 +491,7 @@ static enum dma_status ioat_dma_is_complete(struct dma_chan *chan,
        last_complete = ioat_chan->completed_cookie;
 
        if (done)
-               *done= last_complete;
+               *done = last_complete;
        if (used)
                *used = last_used;
 
@@ -495,7 +505,7 @@ static enum dma_status ioat_dma_is_complete(struct dma_chan *chan,
        last_complete = ioat_chan->completed_cookie;
 
        if (done)
-               *done= last_complete;
+               *done = last_complete;
        if (used)
                *used = last_used;
 
@@ -538,13 +548,13 @@ static irqreturn_t ioat_do_interrupt(int irq, void *data)
 
        attnstatus = readl(instance->reg_base + IOAT_ATTNSTATUS_OFFSET);
 
-       printk(KERN_ERR "ioatdma error: interrupt! status %lx\n", attnstatus);
+       printk(KERN_ERR "ioatdma: interrupt! status %lx\n", attnstatus);
 
        writeb(intrctrl, instance->reg_base + IOAT_INTRCTRL_OFFSET);
        return IRQ_HANDLED;
 }
 
-static void ioat_start_null_desc(struct ioat_dma_chan *ioat_chan)
+static void ioat_dma_start_null_desc(struct ioat_dma_chan *ioat_chan)
 {
        struct ioat_desc_sw *desc;
 
@@ -608,9 +618,11 @@ static int ioat_self_test(struct ioat_device *device)
 
        /* Start copy, using first DMA channel */
        dma_chan = container_of(device->common.channels.next,
-                               struct dma_chan,
-                               device_node);
+                               struct dma_chan,
+                               device_node);
        if (ioat_dma_alloc_chan_resources(dma_chan) < 1) {
+               dev_err(&device->pdev->dev,
+                       "selftest cannot allocate chan resource\n");
                err = -ENODEV;
                goto out;
        }
@@ -628,12 +640,14 @@ static int ioat_self_test(struct ioat_device *device)
        msleep(1);
 
        if (ioat_dma_is_complete(dma_chan, cookie, NULL, NULL) != DMA_SUCCESS) {
-               printk(KERN_ERR "ioatdma: Self-test copy timed out, disabling\n");
+               dev_err(&device->pdev->dev,
+                       "ioatdma: Self-test copy timed out, disabling\n");
                err = -ENODEV;
                goto free_resources;
        }
        if (memcmp(src, dest, IOAT_TEST_SIZE)) {
-               printk(KERN_ERR "ioatdma: Self-test copy failed compare, disabling\n");
+               dev_err(&device->pdev->dev,
+                       "ioatdma: Self-test copy failed compare, disabling\n");
                err = -ENODEV;
                goto free_resources;
        }
@@ -647,7 +661,7 @@ out:
 }
 
 static int __devinit ioat_probe(struct pci_dev *pdev,
-                                const struct pci_device_id *ent)
+                               const struct pci_device_id *ent)
 {
        int err;
        unsigned long mmio_start, mmio_len;
@@ -691,7 +705,9 @@ static int __devinit ioat_probe(struct pci_dev *pdev,
                goto err_dma_pool;
        }
 
-       device->completion_pool = pci_pool_create("completion_pool", pdev, sizeof(u64), SMP_CACHE_BYTES, SMP_CACHE_BYTES);
+       device->completion_pool = pci_pool_create("completion_pool", pdev,
+                                                 sizeof(u64), SMP_CACHE_BYTES,
+                                                 SMP_CACHE_BYTES);
        if (!device->completion_pool) {
                err = -ENOMEM;
                goto err_completion_pool;
@@ -713,22 +729,26 @@ static int __devinit ioat_probe(struct pci_dev *pdev,
 
        device->reg_base = reg_base;
 
-       writeb(IOAT_INTRCTRL_MASTER_INT_EN, device->reg_base + IOAT_INTRCTRL_OFFSET);
+       writeb(IOAT_INTRCTRL_MASTER_INT_EN,
+              device->reg_base + IOAT_INTRCTRL_OFFSET);
        pci_set_master(pdev);
 
        INIT_LIST_HEAD(&device->common.channels);
-       enumerate_dma_channels(device);
+       ioat_dma_enumerate_channels(device);
 
        dma_cap_set(DMA_MEMCPY, device->common.cap_mask);
-       device->common.device_alloc_chan_resources = ioat_dma_alloc_chan_resources;
-       device->common.device_free_chan_resources = ioat_dma_free_chan_resources;
+       device->common.device_alloc_chan_resources =
+                                               ioat_dma_alloc_chan_resources;
+       device->common.device_free_chan_resources =
+                                               ioat_dma_free_chan_resources;
        device->common.device_prep_dma_memcpy = ioat_dma_prep_memcpy;
        device->common.device_is_tx_complete = ioat_dma_is_complete;
        device->common.device_issue_pending = ioat_dma_memcpy_issue_pending;
        device->common.device_dependency_added = ioat_dma_dependency_added;
        device->common.dev = &pdev->dev;
-       printk(KERN_INFO "Intel(R) I/OAT DMA Engine found, %d channels\n",
-               device->common.chancnt);
+       printk(KERN_INFO
+                "ioatdma: Intel(R) I/OAT DMA Engine found, %d channels\n",
+                device->common.chancnt);
 
        err = ioat_self_test(device);
        if (err)
@@ -754,7 +774,8 @@ err_set_dma_mask:
        pci_disable_device(pdev);
 err_enable_device:
 
-       printk(KERN_ERR "Intel(R) I/OAT DMA Engine initialization failed\n");
+       printk(KERN_INFO
+               "ioatdma: Intel(R) I/OAT DMA Engine initialization failed\n");
 
        return err;
 }
@@ -786,7 +807,8 @@ static void __devexit ioat_remove(struct pci_dev *pdev)
        iounmap(device->reg_base);
        pci_release_regions(pdev);
        pci_disable_device(pdev);
-       list_for_each_entry_safe(chan, _chan, &device->common.channels, device_node) {
+       list_for_each_entry_safe(chan, _chan,
+                                &device->common.channels, device_node) {
                ioat_chan = to_ioat_chan(chan);
                list_del(&chan->device_node);
                kfree(ioat_chan);