lkml.org 
[lkml]   [2013]   [Oct]   [3]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [PATCH 0/7] mmc: core: cleanup and locking patches description
From
Ping?

Has anyone had a chance to review/test this series and/or be willing
to do so this week?

grundler@firesword:~$ pwclient list -w 'grundler@chromium.org' -p LKML
Patches submitted by Grant Grundler <grundler@chromium.org>:
ID State Name
-- ----- ----
...
2950961 New [1/7] mmc: core: rename "data" to saved_areq
2951081 New [2/7] mmc: core: rename local var err to saved_err
2950981 New [3/7] mmc: core: restructure error handling for start req
2951001 New [4/7] mmc: core: use common code path to return error
2951061 New [5/7] mmc: core: handling polling more gracefully
2951041 New [6/7] mmc: core: protect references to host->areq
with host->lock
2951021 New [7/7] mmc: core: mmc_start_req is a misnomer ->
mmc_process_areq


To remind, this is the fio command line that results in mmcqd hangs on
the exynos 5250 system (uses dw_mmc storage controller):

$FIO --name=short_randwrite --size=2G --time_based --runtime=90 \
--readwrite=randwrite --numjobs=2 --bs=4k --norandommap
--ioengine=psync \
--direct=0 --invalidate=1 --filename=/dev/mmcblk0p5

thanks,
grant

On Thu, Sep 26, 2013 at 12:57 PM, Grant Grundler <grundler@chromium.org> wrote:
> Argh...too much wordsmithing...
>
> On Thu, Sep 26, 2013 at 12:22 PM, Grant Grundler <grundler@chromium.org> wrote:
>> Following 7 patches are mostly cleanup with one key patch around host->areq
>> locking. The host->areq locking problem description is here:
>> http://www.spinics.net/lists/linux-mmc/msg21644.html
>>
>> I do believe this preposed host->areq locking patch is a complete fix.
>
> ... is NOT a complete fix.
>
>> But it appears to fix the problem and is better than nothing.
>
> Still true.
>
> cheers,
> grant
>
>>
>> This patch sequence applies clean to linus' 3.12-rc2 branch and only compile
>> tested in this form. This is a forward port (and cleanup) of the
>> same patch I've been testing on ChromeOS-3.4 tree using Exynos5250 chipset.
>>
>> cheers,
>> grant
>>
>> 0000-mmc-core-cleanups-and-locking-description (this email)
>> 0001-mmc-core-rename-data-to-saved_areq.patch
>> 0002-mmc-core-rename-local-var-err-to-saved_err.patch
>> 0003-mmc-core-restructure-error-handling-for-start-req.patch
>> 0004-mmc-core-use-common-code-path-to-return-error.patch
>> 0005-mmc-core-handling-polling-more-gracefully.patch
>> 0006-mmc-core-protect-references-to-host-areq-with-host-l.patch
>> 0007-mmc-core-mmc_start_req-is-a-misnomer-mmc_process_are.patch
>>
>>
>> drivers/mmc/card/block.c | 8 ++--
>> drivers/mmc/card/mmc_test.c | 4 +-
>> drivers/mmc/core/core.c | 103 +++++++++++++++++++++++++-------------------
>> include/linux/mmc/core.h | 2 +-
>> 4 files changed, 66 insertions(+), 51 deletions(-)


\
 
 \ /
  Last update: 2013-10-03 23:01    [W:0.088 / U:0.124 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site