* Re: usb-storage: Set virt_boundary_mask to avoid SG overflows
@ 2019-11-04 18:20 Daniel Walker
2019-11-04 18:45 ` Alan Stern
0 siblings, 1 reply; 3+ messages in thread
From: Daniel Walker @ 2019-11-04 18:20 UTC (permalink / raw)
To: Jens Axboe, Alan Stern, Greg Kroah-Hartman
Cc: linux-usb, stable, xe-linux-external
Hi,
This is a stable defect report.
We're tracking v4.9 stable for some of our products (i.e. Cisco Systems, Inc.)
We noticed a speed degradation of roughly %30 on writes to a /dev/sdaX device
over USB (no file system). We bisected the issue to this commit from Alan Stern.
We also found a prior report of speed degradation on NTFS,
https://lore.kernel.org/linux-usb/Pine.LNX.4.44L0.1908291030400.1306-100000@iolanthe.rowland.org/T/
We have the patch reverted in our v4.9 tree on top of stable. It seems Alan was
planning to remove these lines. If the lines are planned to be removed is there
an reason why they haven't been ?
Thanks,
Daniel
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: usb-storage: Set virt_boundary_mask to avoid SG overflows
2019-11-04 18:20 usb-storage: Set virt_boundary_mask to avoid SG overflows Daniel Walker
@ 2019-11-04 18:45 ` Alan Stern
2019-11-04 19:01 ` Daniel Walker
0 siblings, 1 reply; 3+ messages in thread
From: Alan Stern @ 2019-11-04 18:45 UTC (permalink / raw)
To: Daniel Walker
Cc: Jens Axboe, Greg Kroah-Hartman, linux-usb, stable,
xe-linux-external
On Mon, 4 Nov 2019, Daniel Walker wrote:
> Hi,
>
> This is a stable defect report.
>
> We're tracking v4.9 stable for some of our products (i.e. Cisco Systems, Inc.)
> We noticed a speed degradation of roughly %30 on writes to a /dev/sdaX device
> over USB (no file system). We bisected the issue to this commit from Alan Stern.
> We also found a prior report of speed degradation on NTFS,
>
> https://lore.kernel.org/linux-usb/Pine.LNX.4.44L0.1908291030400.1306-100000@iolanthe.rowland.org/T/
>
> We have the patch reverted in our v4.9 tree on top of stable. It seems Alan was
> planning to remove these lines. If the lines are planned to be removed is there
> an reason why they haven't been ?
See https://marc.info/?l=linux-usb&m=157167288816325&w=2
Alan Stern
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: usb-storage: Set virt_boundary_mask to avoid SG overflows
2019-11-04 18:45 ` Alan Stern
@ 2019-11-04 19:01 ` Daniel Walker
0 siblings, 0 replies; 3+ messages in thread
From: Daniel Walker @ 2019-11-04 19:01 UTC (permalink / raw)
To: Alan Stern
Cc: Jens Axboe, Greg Kroah-Hartman, linux-usb, stable,
xe-linux-external
On Mon, Nov 04, 2019 at 01:45:11PM -0500, Alan Stern wrote:
> On Mon, 4 Nov 2019, Daniel Walker wrote:
>
> > Hi,
> >
> > This is a stable defect report.
> >
> > We're tracking v4.9 stable for some of our products (i.e. Cisco Systems, Inc.)
> > We noticed a speed degradation of roughly %30 on writes to a /dev/sdaX device
> > over USB (no file system). We bisected the issue to this commit from Alan Stern.
> > We also found a prior report of speed degradation on NTFS,
> >
> > https://lore.kernel.org/linux-usb/Pine.LNX.4.44L0.1908291030400.1306-100000@iolanthe.rowland.org/T/
> >
> > We have the patch reverted in our v4.9 tree on top of stable. It seems Alan was
> > planning to remove these lines. If the lines are planned to be removed is there
> > an reason why they haven't been ?
>
> See https://marc.info/?l=linux-usb&m=157167288816325&w=2
>
Ok .. Thanks.
Daniel
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2019-11-04 19:01 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2019-11-04 18:20 usb-storage: Set virt_boundary_mask to avoid SG overflows Daniel Walker
2019-11-04 18:45 ` Alan Stern
2019-11-04 19:01 ` Daniel Walker
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).