On 12/08/2017 06:21 PM, Bart Van Assche wrote: > The scsi_sd_probe_domain allows to wait until all disk-probing > activity has finished system-wide. This slows down SCSI host removal > that occurs concurrently with SCSI disk probing because sd_remove() > waits on scsi_sd_probe_domain. Additionally, since each function that > waits on scsi_sd_probe_domain specifies for which disk to wait until > probing has finished, replace waiting on scsi_sd_probe_domain by > waiting until probing for a specific disk has finished. Introduce a > .sync() function pointer in struct scsi_driver to make it possible > for the SCSI power management code to wait until probing of a > specific disk has finished. > > Signed-off-by: Bart Van Assche <bart.vanassche@xxxxxxx> > Cc: Christoph Hellwig <hch@xxxxxx> > Cc: Hannes Reinecke <hare@xxxxxxxx> > Cc: Johannes Thumshirn <jthumshirn@xxxxxxx> > --- > drivers/scsi/scsi.c | 5 ----- > drivers/scsi/scsi_pm.c | 6 ++++-- > drivers/scsi/scsi_priv.h | 1 - > drivers/scsi/sd.c | 26 +++++++++++++++++++++----- > drivers/scsi/sd.h | 1 + > include/scsi/scsi_driver.h | 1 + > 6 files changed, 27 insertions(+), 13 deletions(-) > You know what, I have been working on a similar patch for quite some time now; however, I've taken the simpler approach of not using async_synchronize_full_domain() but rather async_synchronize_cookie(), which makes for a simpler patch :-) However, in doing so I have encountered several issues which have been exposed by that; the most trivial one being that del_gendisk() doesn't check if GENHD_FL_UP is set, so it'll crash if sd_remove is called before sd_async_probe() is run. The other one is an imbalance between sd_probe and sd_remove; when sd_probe_async() is called _after_ scsi_device_remove_device() (which it will as the synchronization is only after device_del() has been called) it will trip over non-existent sysfs directories in add_disk(). So one need to short-circuit sd_probe_async() for devices in SDEV_CANCEL or SDEV_DEL. However, I'm still waiting for a final confirmation on the latter issue, hence I haven't posted the patchset. If there's interest I can post them, though. Cheers, Hannes -- Dr. Hannes Reinecke Teamlead Storage & Networking hare@xxxxxxxx +49 911 74053 688 SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg GF: F. Imendörffer, J. Smithard, J. Guild, D. Upmanyu, G. Norton HRB 21284 (AG Nürnberg)
