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

Groups > linux.kernel > #1207891 > unrolled thread

[PATCH 0/2] fix regression in direct I/O to pmem devices

Back to article view | Back to linux.kernel

Contents

  [PATCH 0/2] fix regression in direct I/O to pmem devices Jeff Moyer <jmoyer@redhat.com> - 2015-08-14 22:20 +0200
    [PATCH 1/2] dax: fix O_DIRECT I/O to the last block of a blockdev Jeff Moyer <jmoyer@redhat.com> - 2015-08-14 22:20 +0200
      Re: [PATCH 1/2] dax: fix O_DIRECT I/O to the last block of a blockdev Linda Knippers <linda.knippers@hp.com> - 2015-08-14 23:00 +0200
    Re: [PATCH 0/2] fix regression in direct I/O to pmem devices Dan Williams <dan.j.williams@intel.com> - 2015-08-14 22:30 +0200
      Re: [PATCH 0/2] fix regression in direct I/O to pmem devices Dan Williams <dan.j.williams@intel.com> - 2015-08-14 22:30 +0200
        Re: [PATCH 0/2] fix regression in direct I/O to pmem devices Jeff Moyer <jmoyer@redhat.com> - 2015-08-14 22:30 +0200

#1207891 — [PATCH 0/2] fix regression in direct I/O to pmem devices

Linda Knippers noticed that commit bbab37ddc20b (block: Add support
for DAX reads/writes to block devices) caused a regression in mkfs.xfs.
Further investigation also uncovered issues related to misaligned
partitions.  This patch set addresses the two issues.

[PATCH 1/2] dax: fix O_DIRECT I/O to the last block of a blockdev
[PATCH 2/2] blockdev: don't set S_DAX for misaligned partitions

 fs/block_dev.c | 7 +++++++
 fs/dax.c       | 3 ++-
 2 files changed, 9 insertions(+), 1 deletion(-)

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [next] | [standalone]

#1207892 — [PATCH 1/2] dax: fix O_DIRECT I/O to the last block of a blockdev

commit bbab37ddc20b (block: Add support for DAX reads/writes to
block devices) caused a regression in mkfs.xfs.  That utility
sets the block size of the device to the logical block size
using the BLKBSZSET ioctl, and then issues a single sector read
from the last sector of the device.  This results in the dax_io
code trying to do a page-sized read from 512 bytes from the end
of the device.  The result is -ERANGE being returned to userspace.

The fix is to align the block to the page size before calling
get_block.

Thanks to willy for simplifying my original patch.

Signed-off-by: Jeff Moyer <jmoyer@redhat.com>
---
 fs/dax.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/dax.c b/fs/dax.c
index a7f77e1..ef35a20 100644
--- a/fs/dax.c
+++ b/fs/dax.c
@@ -116,7 +116,8 @@ static ssize_t dax_io(struct inode *inode, struct iov_iter *iter,
 		unsigned len;
 		if (pos == max) {
 			unsigned blkbits = inode->i_blkbits;
-			sector_t block = pos >> blkbits;
+			long page = pos >> PAGE_SHIFT;
+			sector_t block = page << (PAGE_SHIFT - blkbits);
 			unsigned first = pos - (block << blkbits);
 			long size;
 
-- 
1.8.3.1

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1207902 — Re: [PATCH 1/2] dax: fix O_DIRECT I/O to the last block of a blockdev

On 8/14/2015 4:15 PM, Jeff Moyer wrote:
> commit bbab37ddc20b (block: Add support for DAX reads/writes to
> block devices) caused a regression in mkfs.xfs.  That utility
> sets the block size of the device to the logical block size
> using the BLKBSZSET ioctl, and then issues a single sector read
> from the last sector of the device.  This results in the dax_io
> code trying to do a page-sized read from 512 bytes from the end
> of the device.  The result is -ERANGE being returned to userspace.
> 
> The fix is to align the block to the page size before calling
> get_block.
> 
> Thanks to willy for simplifying my original patch.
> 
> Signed-off-by: Jeff Moyer <jmoyer@redhat.com>

Tested-by:  Linda Knippers <linda.knippers@hp.com>

> ---
>  fs/dax.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/dax.c b/fs/dax.c
> index a7f77e1..ef35a20 100644
> --- a/fs/dax.c
> +++ b/fs/dax.c
> @@ -116,7 +116,8 @@ static ssize_t dax_io(struct inode *inode, struct iov_iter *iter,
>  		unsigned len;
>  		if (pos == max) {
>  			unsigned blkbits = inode->i_blkbits;
> -			sector_t block = pos >> blkbits;
> +			long page = pos >> PAGE_SHIFT;
> +			sector_t block = page << (PAGE_SHIFT - blkbits);
>  			unsigned first = pos - (block << blkbits);
>  			long size;
>  
> 

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1207894

On Fri, Aug 14, 2015 at 1:15 PM, Jeff Moyer <jmoyer@redhat.com> wrote:
> Linda Knippers noticed that commit bbab37ddc20b (block: Add support
> for DAX reads/writes to block devices) caused a regression in mkfs.xfs.
> Further investigation also uncovered issues related to misaligned
> partitions.  This patch set addresses the two issues.
>
> [PATCH 1/2] dax: fix O_DIRECT I/O to the last block of a blockdev
> [PATCH 2/2] blockdev: don't set S_DAX for misaligned partitions

Should these be "Cc: <stable@vger.kernel.org>" given the regression
fix is applicable back to 4.0?
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1207895

On Fri, Aug 14, 2015 at 1:22 PM, Dan Williams <dan.j.williams@intel.com> wrote:
> On Fri, Aug 14, 2015 at 1:15 PM, Jeff Moyer <jmoyer@redhat.com> wrote:
>> Linda Knippers noticed that commit bbab37ddc20b (block: Add support
>> for DAX reads/writes to block devices) caused a regression in mkfs.xfs.
>> Further investigation also uncovered issues related to misaligned
>> partitions.  This patch set addresses the two issues.
>>
>> [PATCH 1/2] dax: fix O_DIRECT I/O to the last block of a blockdev
>> [PATCH 2/2] blockdev: don't set S_DAX for misaligned partitions
>
> Should these be "Cc: <stable@vger.kernel.org>" given the regression
> fix is applicable back to 4.0?

Sorry, regression *since* 4.0.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [next] | [standalone]

#1207896

Dan Williams <dan.j.williams@intel.com> writes:

> On Fri, Aug 14, 2015 at 1:22 PM, Dan Williams <dan.j.williams@intel.com> wrote:
>> On Fri, Aug 14, 2015 at 1:15 PM, Jeff Moyer <jmoyer@redhat.com> wrote:
>>> Linda Knippers noticed that commit bbab37ddc20b (block: Add support
>>> for DAX reads/writes to block devices) caused a regression in mkfs.xfs.
>>> Further investigation also uncovered issues related to misaligned
>>> partitions.  This patch set addresses the two issues.
>>>
>>> [PATCH 1/2] dax: fix O_DIRECT I/O to the last block of a blockdev
>>> [PATCH 2/2] blockdev: don't set S_DAX for misaligned partitions
>>
>> Should these be "Cc: <stable@vger.kernel.org>" given the regression
>> fix is applicable back to 4.0?
>
> Sorry, regression *since* 4.0.

$ git describe --contains bbab37ddc20b
v4.2-rc1~2^2~4

Looks like this was not sprung on the public yet.

-Jeff
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web