Re: [PATCH] block: file-posix: Fail unmap with NO_FALLBACK on block device

Am 16.06.2020 um 22:01 hat Nir Soffer geschrieben: > On Tue, Jun 16, 2020 at 8:39 PM Nir Soffer wrote: > > > > On Tue, Jun 16, 2020 at 6:32 PM Kevin Wolf wrote: > > > > > > Am 15.06.2020 um 21:32 hat Nir Soffer geschrieben: > > > > We can zero 2.3 g/s: > > > > > > > > # time blkdiscard -z test-lv

Re: [PATCH] block: file-posix: Fail unmap with NO_FALLBACK on block device

On 6/16/20 10:32 AM, Kevin Wolf wrote: Am 15.06.2020 um 21:32 hat Nir Soffer geschrieben: We can zero 2.3 g/s: # time blkdiscard -z test-lv real 0m43.902s user 0m0.002s sys 0m0.130s We can write 445m/s: # dd if=/dev/zero bs=2M count=51200 of=test-lv oflag=direct conv=fsync 107374182400 byt

Re: [PATCH] block: file-posix: Fail unmap with NO_FALLBACK on block device

On Tue, Jun 16, 2020 at 8:39 PM Nir Soffer wrote: > > On Tue, Jun 16, 2020 at 6:32 PM Kevin Wolf wrote: > > > > Am 15.06.2020 um 21:32 hat Nir Soffer geschrieben: > > > We can zero 2.3 g/s: > > > > > > # time blkdiscard -z test-lv > > > > > > real 0m43.902s > > > user 0m0.002s > > > sys 0m0.130s

Re: [PATCH] block: file-posix: Fail unmap with NO_FALLBACK on block device

On Tue, Jun 16, 2020 at 6:32 PM Kevin Wolf wrote: > > Am 15.06.2020 um 21:32 hat Nir Soffer geschrieben: > > We can zero 2.3 g/s: > > > > # time blkdiscard -z test-lv > > > > real 0m43.902s > > user 0m0.002s > > sys 0m0.130s > > > We can write 445m/s: > > > > # dd if=/dev/zero bs=2M count=51200 of

Re: [PATCH] block: file-posix: Fail unmap with NO_FALLBACK on block device

Am 15.06.2020 um 21:32 hat Nir Soffer geschrieben: > We can zero 2.3 g/s: > > # time blkdiscard -z test-lv > > real 0m43.902s > user 0m0.002s > sys 0m0.130s > We can write 445m/s: > > # dd if=/dev/zero bs=2M count=51200 of=test-lv oflag=direct conv=fsync > 107374182400 bytes (107 GB, 100 GiB) c

Re: [PATCH] block: file-posix: Fail unmap with NO_FALLBACK on block device

On Sat, Jun 13, 2020 at 8:08 PM Nir Soffer wrote: > > Punching holes on block device uses blkdev_issue_zeroout() with > BLKDEV_ZERO_NOFALLBACK but there is no guarantee that this is fast > enough for pre-zeroing an entire device. > > Zeroing block device can be slow as writing zeroes or 100 times

[PATCH] block: file-posix: Fail unmap with NO_FALLBACK on block device

Punching holes on block device uses blkdev_issue_zeroout() with BLKDEV_ZERO_NOFALLBACK but there is no guarantee that this is fast enough for pre-zeroing an entire device. Zeroing block device can be slow as writing zeroes or 100 times faster, depending on the storage. There is no way to tell if z