| Message ID | 20180704085356.10888-1-jthumshirn@suse.de (mailing list archive) |
|---|---|
| State | Accepted |
| Headers | show |
Johannes, > It has been more than one year since we tried to change the default > from legacy to multi queue in SCSI with commit c279bd9e406 ("scsi: > default to scsi-mq"). But due to issues with suspend/resume and > performance problems it had been reverted again with commit > cbe7dfa26eee ("Revert "scsi: default to scsi-mq""). > > In the meantime there have been a substantial amount of performance > improvements and suspend/resume got fixed as well, thus we can > re-enable scsi-mq without a significant performance penalty. Applied to 4.19/scsi-queue, thanks!
On 11/07/18 05:52, Martin K. Petersen wrote: > > Johannes, > >> It has been more than one year since we tried to change the default >> from legacy to multi queue in SCSI with commit c279bd9e406 ("scsi: >> default to scsi-mq"). But due to issues with suspend/resume and >> performance problems it had been reverted again with commit >> cbe7dfa26eee ("Revert "scsi: default to scsi-mq""). >> >> In the meantime there have been a substantial amount of performance >> improvements and suspend/resume got fixed as well, thus we can >> re-enable scsi-mq without a significant performance penalty. > > Applied to 4.19/scsi-queue, thanks! > But there is no runtime pm support, is there? I tried enabling runtime PM for UFS SCSI devices with scsi-mq and it didn't work, which is a regression.
On Wed, Jul 11, 2018 at 4:31 PM, Adrian Hunter <adrian.hunter@intel.com> wrote: > On 11/07/18 05:52, Martin K. Petersen wrote: >> >> Johannes, >> >>> It has been more than one year since we tried to change the default >>> from legacy to multi queue in SCSI with commit c279bd9e406 ("scsi: >>> default to scsi-mq"). But due to issues with suspend/resume and >>> performance problems it had been reverted again with commit >>> cbe7dfa26eee ("Revert "scsi: default to scsi-mq""). >>> >>> In the meantime there have been a substantial amount of performance >>> improvements and suspend/resume got fixed as well, thus we can >>> re-enable scsi-mq without a significant performance penalty. >> >> Applied to 4.19/scsi-queue, thanks! >> > > But there is no runtime pm support, is there? I tried enabling runtime PM > for UFS SCSI devices with scsi-mq and it didn't work, which is a regression. I just posted one RFC patchset for supporting runtime PM on scsi_mq, and looks it works in my test on usb-storage. https://marc.info/?l=linux-block&m=153132657922233&w=2 thanks, Ming Lei
On 11/07/18 19:35, Ming Lei wrote: > On Wed, Jul 11, 2018 at 4:31 PM, Adrian Hunter <adrian.hunter@intel.com> wrote: >> On 11/07/18 05:52, Martin K. Petersen wrote: >>> >>> Johannes, >>> >>>> It has been more than one year since we tried to change the default >>>> from legacy to multi queue in SCSI with commit c279bd9e406 ("scsi: >>>> default to scsi-mq"). But due to issues with suspend/resume and >>>> performance problems it had been reverted again with commit >>>> cbe7dfa26eee ("Revert "scsi: default to scsi-mq""). >>>> >>>> In the meantime there have been a substantial amount of performance >>>> improvements and suspend/resume got fixed as well, thus we can >>>> re-enable scsi-mq without a significant performance penalty. >>> >>> Applied to 4.19/scsi-queue, thanks! >>> >> >> But there is no runtime pm support, is there? I tried enabling runtime PM >> for UFS SCSI devices with scsi-mq and it didn't work, which is a regression. > > I just posted one RFC patchset for supporting runtime PM on scsi_mq, and > looks it works in my test on usb-storage. > > https://marc.info/?l=linux-block&m=153132657922233&w=2 Cool, please cc me on the next revision of your patch set.
diff --git a/drivers/scsi/Kconfig b/drivers/scsi/Kconfig index 35c909bbf8ba..bd115bab162e 100644 --- a/drivers/scsi/Kconfig +++ b/drivers/scsi/Kconfig @@ -49,6 +49,7 @@ config SCSI_NETLINK config SCSI_MQ_DEFAULT bool "SCSI: use blk-mq I/O path by default" + default y depends on SCSI ---help--- This option enables the new blk-mq based I/O path for SCSI
It has been more than one year since we tried to change the default from legacy to multi queue in SCSI with commit c279bd9e406 ("scsi: default to scsi-mq"). But due to issues with suspend/resume and performance problems it had been reverted again with commit cbe7dfa26eee ("Revert "scsi: default to scsi-mq""). In the meantime there have been a substantial amount of performance improvements and suspend/resume got fixed as well, thus we can re-enable scsi-mq without a significant performance penalty. Signed-off-by: Johannes Thumshirn <jthumshirn@suse.de> Reviewed-by: Hannes Reinecke <hare@suse.com> Reviewed-by: Ming Lei <ming.lei@redhat.com> Acked-by: John Garry <john.garry@huawei.com> --- Changes since RFC: - Improve Changelog as Bart suggested - Collect Acks and Reviews --- drivers/scsi/Kconfig | 1 + 1 file changed, 1 insertion(+)