| Message ID | 956CEF49-A326-4F68-BCB3-350C4BF3BAA8@fb.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [GIT,PULL] md-next 2023-09-28 | expand |
On 9/28/23 2:58 PM, Song Liu wrote: > Hi Jens, > > Please consider pulling the following changes for md-next on top of your > for-6.7/block branch. > > Major changes in these patches are: > > 1. Make rdev add/remove independent from daemon thread, by Yu Kuai; > 2. Refactor code around quiesce() and mddev_suspend(), by Yu Kuai. Changes looks fine to me, but this patch: https://git.kernel.org/pub/scm/linux/kernel/git/song/md.git/commit/?h=md-next&id=b71fe4ac7531d67e6fc8c287cbcb2b176aa93833 is referencing a commit that doesn't exist: "After commit 4d27e927344a ("md: don't quiesce in mddev_suspend()")," which I think should be: b39f35ebe86d ("md: don't quiesce in mddev_suspend()") where is this other sha from?
> On Sep 28, 2023, at 11:02 PM, Jens Axboe <axboe@kernel.dk> wrote: > > On 9/28/23 2:58 PM, Song Liu wrote: >> Hi Jens, >> >> Please consider pulling the following changes for md-next on top of your >> for-6.7/block branch. >> >> Major changes in these patches are: >> >> 1. Make rdev add/remove independent from daemon thread, by Yu Kuai; >> 2. Refactor code around quiesce() and mddev_suspend(), by Yu Kuai. > > Changes looks fine to me, but this patch: > > https://git.kernel.org/pub/scm/linux/kernel/git/song/md.git/commit/?h=md-next&id=b71fe4ac7531d67e6fc8c287cbcb2b176aa93833 > > is referencing a commit that doesn't exist: > > "After commit 4d27e927344a ("md: don't quiesce in mddev_suspend()")," > > which I think should be: > > b39f35ebe86d ("md: don't quiesce in mddev_suspend()") > > where is this other sha from? The other sha was from a previous md-next that got rebased later. I guess Kuai didn't catch it because he had the old sha in his local git cache. I should have caught it, but missed it because it was not behind a Fixed tag (still my fault). I recently improved my process to only do rebase when necessary (I used to rebase too often). Hopefully this will prevent sha mismatch in the future. However, to fix this one, I guess I have no options but rebase the branch? I will fix it later today and resend the pull request. I am really sorry for having the same issue again. Thanks for catching it. Song
On 9/29/23 6:12 AM, Song Liu wrote: > > >> On Sep 28, 2023, at 11:02 PM, Jens Axboe <axboe@kernel.dk> wrote: >> >> On 9/28/23 2:58 PM, Song Liu wrote: >>> Hi Jens, >>> >>> Please consider pulling the following changes for md-next on top of your >>> for-6.7/block branch. >>> >>> Major changes in these patches are: >>> >>> 1. Make rdev add/remove independent from daemon thread, by Yu Kuai; >>> 2. Refactor code around quiesce() and mddev_suspend(), by Yu Kuai. >> >> Changes looks fine to me, but this patch: >> >> https://git.kernel.org/pub/scm/linux/kernel/git/song/md.git/commit/?h=md-next&id=b71fe4ac7531d67e6fc8c287cbcb2b176aa93833 >> >> is referencing a commit that doesn't exist: >> >> "After commit 4d27e927344a ("md: don't quiesce in mddev_suspend()")," >> >> which I think should be: >> >> b39f35ebe86d ("md: don't quiesce in mddev_suspend()") >> >> where is this other sha from? > > The other sha was from a previous md-next that got rebased later. > I guess Kuai didn't catch it because he had the old sha in his > local git cache. I should have caught it, but missed it because > it was not behind a Fixed tag (still my fault). Makes sense. It's not a huge deal, and I would not rebase your tree for it. Mostly just curious where it came from. > I recently improved my process to only do rebase when necessary > (I used to rebase too often). Hopefully this will prevent sha > mismatch in the future. However, to fix this one, I guess I have > no options but rebase the branch? I will fix it later today and > resend the pull request. I did notice that the tree looks much better now, rather than have all brand new commits. Thanks for doing that! So I say we keep it as-is and I pull it - it's easy enough to find the real commit just based on the subject.
> On Sep 29, 2023, at 5:16 AM, Jens Axboe <axboe@kernel.dk> wrote: > > On 9/29/23 6:12 AM, Song Liu wrote: >> >> >>> On Sep 28, 2023, at 11:02 PM, Jens Axboe <axboe@kernel.dk> wrote: >>> >>> On 9/28/23 2:58 PM, Song Liu wrote: >>>> Hi Jens, >>>> >>>> Please consider pulling the following changes for md-next on top of your >>>> for-6.7/block branch. >>>> >>>> Major changes in these patches are: >>>> >>>> 1. Make rdev add/remove independent from daemon thread, by Yu Kuai; >>>> 2. Refactor code around quiesce() and mddev_suspend(), by Yu Kuai. >>> >>> Changes looks fine to me, but this patch: >>> >>> https://git.kernel.org/pub/scm/linux/kernel/git/song/md.git/commit/?h=md-next&id=b71fe4ac7531d67e6fc8c287cbcb2b176aa93833 >>> >>> is referencing a commit that doesn't exist: >>> >>> "After commit 4d27e927344a ("md: don't quiesce in mddev_suspend()")," >>> >>> which I think should be: >>> >>> b39f35ebe86d ("md: don't quiesce in mddev_suspend()") >>> >>> where is this other sha from? >> >> The other sha was from a previous md-next that got rebased later. >> I guess Kuai didn't catch it because he had the old sha in his >> local git cache. I should have caught it, but missed it because >> it was not behind a Fixed tag (still my fault). > > Makes sense. It's not a huge deal, and I would not rebase your tree for > it. Mostly just curious where it came from. Got it. Thanks for the suggestion. > >> I recently improved my process to only do rebase when necessary >> (I used to rebase too often). Hopefully this will prevent sha >> mismatch in the future. However, to fix this one, I guess I have >> no options but rebase the branch? I will fix it later today and >> resend the pull request. > > I did notice that the tree looks much better now, rather than have all > brand new commits. Thanks for doing that! So I say we keep it as-is and > I pull it - it's easy enough to find the real commit just based on the > subject. Yeah, the real commit is easy to find. Song
Hi Jens, Please consider pulling the following changes for md-next on top of your for-6.7/block branch. Major changes in these patches are: 1. Make rdev add/remove independent from daemon thread, by Yu Kuai; 2. Refactor code around quiesce() and mddev_suspend(), by Yu Kuai. Thanks, Song The following changes since commit d78bfa1346ab1fe04d20aa45a0678d1fc866f37c: block/null_blk: add queue_rqs() support (2023-09-22 08:52:13 -0600) are available in the Git repository at: https://git.kernel.org/pub/scm/linux/kernel/git/song/md.git tags/md-next-20230927 for you to fetch changes up to ceb0416383988dbd5decd6a70141a3507732c160: md: replace deprecated strncpy with memcpy (2023-09-25 14:36:41 -0700) ---------------------------------------------------------------- Justin Stitt (1): md: replace deprecated strncpy with memcpy Kees Cook (1): md/md-linear: Annotate struct linear_conf with __counted_by Yu Kuai (14): md: use separate work_struct for md_start_sync() md: factor out a helper to choose sync action from md_check_recovery() md: delay choosing sync action to md_start_sync() md: factor out a helper rdev_removeable() from remove_and_add_spares() md: factor out a helper rdev_is_spare() from remove_and_add_spares() md: factor out a helper rdev_addable() from remove_and_add_spares() md: delay remove_and_add_spares() for read only array to md_start_sync() md: initialize 'active_io' while allocating mddev md: initialize 'writes_pending' while allocating mddev md: don't rely on 'mddev->pers' to be set in mddev_suspend() md-bitmap: remove the checking of 'pers->quiesce' from location_store() md-bitmap: suspend array earlier in location_store() md: don't check 'mddev->pers' from suspend_hi_store() md: don't check 'mddev->pers' and 'pers->quiesce' from suspend_lo_store() drivers/md/dm-raid.c | 7 ++- drivers/md/md-bitmap.c | 47 ++++++++---------- drivers/md/md-linear.c | 26 +++++----- drivers/md/md-linear.h | 2 +- drivers/md/md.c | 414 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++--------------------------------------------------------- drivers/md/md.h | 9 ++-- drivers/md/raid1.c | 3 +- drivers/md/raid10.c | 3 -- drivers/md/raid5.c | 3 -- 9 files changed, 313 insertions(+), 201 deletions(-)