Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1734748 > unrolled thread

[PATCH V6 3/3] scsi: Align block queue to dma_get_cache_alignment()

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  [PATCH V6 3/3] scsi: Align block queue to dma_get_cache_alignment() Huacai Chen <chenhc@lemote.com> - 2017-09-19 11:00 +0200
    Re: [PATCH V6 3/3] scsi: Align block queue to  dma_get_cache_alignment() kbuild test robot <lkp@intel.com> - 2017-09-24 05:50 +0200

#1734748 — [PATCH V6 3/3] scsi: Align block queue to dma_get_cache_alignment()

In non-coherent DMA mode, kernel uses cache flushing operations to
maintain I/O coherency, so scsi's block queue should be aligned to
ARCH_DMA_MINALIGN. Otherwise, it will cause data corruption, at least
on MIPS:

        Step 1, dma_map_single
        Step 2, cache_invalidate (no writeback)
        Step 3, dma_from_device
        Step 4, dma_unmap_single

If a DMA buffer and a kernel structure share a same cache line, and if
the kernel structure has dirty data, cache_invalidate (no writeback)
will cause data lost.

Cc: stable@vger.kernel.org
Signed-off-by: Huacai Chen <chenhc@lemote.com>
---
 drivers/scsi/scsi_lib.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/scsi/scsi_lib.c b/drivers/scsi/scsi_lib.c
index 9cf6a80..19abc2e 100644
--- a/drivers/scsi/scsi_lib.c
+++ b/drivers/scsi/scsi_lib.c
@@ -2132,11 +2132,11 @@ void __scsi_init_queue(struct Scsi_Host *shost, struct request_queue *q)
 		q->limits.cluster = 0;
 
 	/*
-	 * set a reasonable default alignment on word boundaries: the
-	 * host and device may alter it using
+	 * set a reasonable default alignment on word/cacheline boundaries:
+	 * the host and device may alter it using
 	 * blk_queue_update_dma_alignment() later.
 	 */
-	blk_queue_dma_alignment(q, 0x03);
+	blk_queue_dma_alignment(q, max(4, dma_get_cache_alignment(dev)) - 1);
 }
 EXPORT_SYMBOL_GPL(__scsi_init_queue);
 
-- 
2.7.0

[toc] | [next] | [standalone]

#1738105 — Re: [PATCH V6 3/3] scsi: Align block queue to dma_get_cache_alignment()

[Multipart message — attachments visible in raw view] — view raw

Hi Huacai,

[auto build test ERROR on linus/master]
[also build test ERROR on v4.14-rc1 next-20170922]
[if your patch is applied to the wrong git tree, please drop us a note to help improve the system]

url:    https://github.com/0day-ci/linux/commits/Huacai-Chen/dma-mapping-Introduce-device_is_coherent-as-a-helper/20170920-204740
config: m68k-sun3_defconfig (attached as .config)
compiler: m68k-linux-gcc (GCC) 4.9.0
reproduce:
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # save the attached .config to linux build tree
        make.cross ARCH=m68k 

All errors (new ones prefixed by >>):

   drivers/scsi/scsi_lib.c: In function '__scsi_init_queue':
>> drivers/scsi/scsi_lib.c:2139:2: error: implicit declaration of function 'dma_get_cache_alignment' [-Werror=implicit-function-declaration]
     blk_queue_dma_alignment(q, max(4, dma_get_cache_alignment(dev)) - 1);
     ^
   cc1: some warnings being treated as errors

vim +/dma_get_cache_alignment +2139 drivers/scsi/scsi_lib.c

  2103	
  2104	void __scsi_init_queue(struct Scsi_Host *shost, struct request_queue *q)
  2105	{
  2106		struct device *dev = shost->dma_dev;
  2107	
  2108		queue_flag_set_unlocked(QUEUE_FLAG_SCSI_PASSTHROUGH, q);
  2109	
  2110		/*
  2111		 * this limit is imposed by hardware restrictions
  2112		 */
  2113		blk_queue_max_segments(q, min_t(unsigned short, shost->sg_tablesize,
  2114						SG_MAX_SEGMENTS));
  2115	
  2116		if (scsi_host_prot_dma(shost)) {
  2117			shost->sg_prot_tablesize =
  2118				min_not_zero(shost->sg_prot_tablesize,
  2119					     (unsigned short)SCSI_MAX_PROT_SG_SEGMENTS);
  2120			BUG_ON(shost->sg_prot_tablesize < shost->sg_tablesize);
  2121			blk_queue_max_integrity_segments(q, shost->sg_prot_tablesize);
  2122		}
  2123	
  2124		blk_queue_max_hw_sectors(q, shost->max_sectors);
  2125		blk_queue_bounce_limit(q, scsi_calculate_bounce_limit(shost));
  2126		blk_queue_segment_boundary(q, shost->dma_boundary);
  2127		dma_set_seg_boundary(dev, shost->dma_boundary);
  2128	
  2129		blk_queue_max_segment_size(q, dma_get_max_seg_size(dev));
  2130	
  2131		if (!shost->use_clustering)
  2132			q->limits.cluster = 0;
  2133	
  2134		/*
  2135		 * set a reasonable default alignment on word/cacheline boundaries:
  2136		 * the host and device may alter it using
  2137		 * blk_queue_update_dma_alignment() later.
  2138		 */
> 2139		blk_queue_dma_alignment(q, max(4, dma_get_cache_alignment(dev)) - 1);
  2140	}
  2141	EXPORT_SYMBOL_GPL(__scsi_init_queue);
  2142	

---
0-DAY kernel test infrastructure                Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all                   Intel Corporation

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web