disable TCQ?

2007-06-28 Thread Michael Tokarev
/sys/block/sdX/device/queue_depth file is read-write with libsata, but is read-only with SCSI controllers (at least with the ones we have -- mostly aic7xxx or mptsas). Googling for "disable tcq" gives results about {dis,en}abling TCQ at boot time, and also mentions ipr (which we don't have). Is

Re: disable TCQ?

2007-06-28 Thread Jeff Garzik
Michael Tokarev wrote: /sys/block/sdX/device/queue_depth file is read-write with libsata, but is read-only with SCSI controllers (at least with the ones we have -- mostly aic7xxx or mptsas). Googling for "disable tcq" gives results about {dis,en}abling TCQ at boot time, and also mentions ipr (w

Some NCQ numbers...

2007-06-28 Thread Michael Tokarev
[Offtopic notice: For the first time I demonstrated some speed testing results on linux-ide mailinglist, as a demonstration how [NT]CQ can help. But later, someone becomes curious and posted that email to lkml, asking for more details. Since that, I become more curious as well, and decided to loo

Re: Some NCQ numbers...

2007-06-28 Thread Michael Tokarev
Michael Tokarev wrote: [] > I'm planning to test several models of SCSI drives. On SCSI front > (or maybe with different drives - I don't know) things are WAY more > interesting wrt TCQ. Difference in results between 1 and 32 threads > goes up to 4 times sometimes!. But I'm a bit stuck with SCSI

[PATCH] CONFIG_SCSI_FD_8xx no longer exists

2007-06-28 Thread Geert Uytterhoeven
CONFIG_SCSI_FD_8xx no longer exists. Apparently it was renamed to CONFIG_SCSI_SEAGATE, but the Makefile was not correctly updated. Signed-off-by: Geert Uytterhoeven <[EMAIL PROTECTED]> diff --git a/drivers/scsi/Makefile b/drivers/scsi/Makefile index b1b6327..eb92b31 100644 --- a/drivers/scsi/Ma

Re: [PATCH] zfcp: Report FCP LUN to SCSI midlayer

2007-06-28 Thread Stefan Richter
Luben Tuikov wrote: > In fact, I'd do > typedef __u8 lun_t[8]; > and then define the macro > #define LUN_TO_U64(_lun) ((unsigned long long) be64_to_cpu(*(__be64 > *)(_lun))) Don't forget __attribute__((aligned(8))) on lun_t then. Some architectures need it. The macro should perhaps be calle

2.6.22-rc6-mm1: multiple definition of `check_reset'

2007-06-28 Thread Adrian Bunk
On Thu, Jun 28, 2007 at 03:43:21AM -0700, Andrew Morton wrote: >... > Changes since 2.6.22-rc4-mm2: >... > git-scsi-misc.patch >... > git trees >... <-- snip --> ... LD drivers/built-in.o drivers/isdn/built-in.o: In function `check_reset': (.text+0xab974): multiple definition of `chec

Re: 2.6.22-rc6-mm1: multiple definition of `check_reset'

2007-06-28 Thread Andrew Morton
On Thu, 28 Jun 2007 21:29:50 +0200 Adrian Bunk <[EMAIL PROTECTED]> wrote: > LD drivers/built-in.o > drivers/isdn/built-in.o: In function `check_reset': (.text+0xab974): > multiple definition of `check_reset' > drivers/scsi/built-in.o:(.data+0xdba8): first defined here > ld: Warning: size of