On Fri, Nov 30, 2007 at 01:17:22PM +0300, Sergei Shtylyov wrote:
Anton Vorontsov wrote:
Remaining question: any preferred name for that property? pio-mode okay?
It's assuming that PIO6 capable bus supports PIO0 as well, thus no mask.
I've already suggested generic. A name simple also
If one disregards the rotational speed and access time advantage that
SAS drives have over SATA, does the SAS interface offer any performance
advantage?
For example, assume a SAS drive and a SATA drive can both sustained
stream 70MB/s. A 16 drive JBOD SAS enclosure with internal SAS expander
On Fri 2007-11-30 10:00:55, Mark Lord wrote:
Pavel Machek wrote:
On Fri 2007-11-30 13:13:44, Alan Cox wrote:
Why does a single spurious interrupt cause it to be shut down? I can
It doesn't.
see if the interrupt is stuck on and keeps interrupting constantly, but
if it's just the
This patch adds 2 new interfaces for request completion:
o blk_end_request() : called without queue lock
o __blk_end_request() : called with queue lock held
Some device drivers call some generic functions below between
end_that_request_{first/chunk} and end_that_request_last().
o
Hello Jens,
The following is the updated patch-set for blk_end_request().
Changes since the last version are only minor updates to catch up
with the base kernel changes.
Do you agree the implementation of blk_end_request()?
If there's no problem, could you merge it to your tree?
Or does it have
First of all, I've yet to see a controller that is really able
to handle multiple requests in parallel. Usually, multiple
I/O threads gets exactly the same summary performance as a
single thread - UNLIKE of linux software raid which clearly
Thats usually true of hardware raid cards as they
Phillip Susi wrote:
Tejun Heo wrote:
Because SFF ATA controller don't have IRQ pending bit. You don't know
whether IRQ is raised or not. Plus, accessing the status register which
clears pending IRQ can be very slow on PATA machines. It has to go
through the PCI and ATA bus and come back.
Enclosure Management via LED
This patch implements Enclosure Management via the LED protocol as specified
in AHCI specification.
Signed-off-by: Kristen Carlson Accardi [EMAIL PROTECTED]
---
This revision makes the change to the comment requested by Mark Lord,
fixes some bugs in the bit shifting
Thanks for the comments.
It was really protocol/bus behaviour differences (if any), between SATA
drives in a SAS environment, vs SAS drives, that I am looking at.
I do know that SATA drives only support a subset of the SCSI commands
and wondered if the SAS drives were more clever in a multi
Alan Cox wrote:
The comment I saw, which I'm trying to verify, mentioned the SATA drives
held the bus or similar longer than SAS ones.
SATA normally uses one link per device so the device side isn't contended
unless you descend into the murky world of port multipliers.
..
And that's where
This patch removes the following functions:
o end_that_request_first()
o end_that_request_chunk()
and stops exporting the functions below:
o end_that_request_last()
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
block/ll_rw_blk.c |
This patch adds a variant of the interface, blk_end_request_callback(),
which has driver callback feature.
There are 2 drivers which need to do special works between
end_that_request_first() and end_that_request_last():
ide-cd and scsi bidi.
For such drivers, blk_end_request_callback() allows it
This patch converts cciss to use blk_end_request().
cciss is a little bit different from normal drivers.
cciss directly calls bio_endio() and disk_stat_add()
when completing request. But those can be replaced with
__end_that_request_first().
After the replacement, request completion procedures
This patch converts normal parts of ide to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/ide/ide-cd.c |6 +++---
drivers/ide/ide-io.c | 17 ++---
2 files changed, 9 insertions(+), 14
This patch converts s390 to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/s390/block/dasd.c |4 +---
drivers/s390/char/tape_block.c |3 +--
2 files changed, 2 insertions(+), 5 deletions(-)
Index:
This patch converts ide-scsi to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/scsi/ide-scsi.c |8
1 files changed, 4 insertions(+), 4 deletions(-)
Index: 2.6.24-rc3-mm2/drivers/scsi/ide-scsi.c
This patch converts sunvdc to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/block/sunvdc.c |5 +
1 files changed, 1 insertion(+), 4 deletions(-)
Index: 2.6.24-rc3-mm2/drivers/block/sunvdc.c
This patch converts ps3disk to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/block/ps3disk.c |6 +-
1 files changed, 1 insertion(+), 5 deletions(-)
Index: 2.6.24-rc3-mm2/drivers/block/ps3disk.c
This patch converts ide-cd (cdrom_newpc_intr()) to use blk_end_request().
ide-cd (cdrom_newpc_intr()) has some tricky behaviors below which
need to use blk_end_request_callback().
Needs to:
1. call post_transform_command() to modify request contents
2. wait completing request until DRQ_STAT
This patch converts cpqarray to use blk_end_request().
cpqarray is a little bit different from normal drivers.
cpqarray directly calls bio_endio() and disk_stat_add()
when completing request. But those can be replaced with
__end_that_request_first().
After the replacement, request completion
This patch converts scsi mid-layer to use blk_end_request().
The comment above scsi_next_command() is not related to this change.
It had originally been there before scsi_next_command() was included
in scsi_finalize_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi
This patch converts i2o_block to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/message/i2o/i2o_block.c |8 ++--
1 files changed, 2 insertions(+), 6 deletions(-)
Index:
This patch converts viocd to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/cdrom/viocd.c |5 +
1 files changed, 1 insertion(+), 4 deletions(-)
Index: 2.6.24-rc3-mm2/drivers/cdrom/viocd.c
This patch converts viodasd to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/block/viodasd.c |5 +
1 files changed, 1 insertion(+), 4 deletions(-)
Index: 2.6.24-rc3-mm2/drivers/block/viodasd.c
This patch converts floppy to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/block/floppy.c |8 +++-
1 files changed, 3 insertions(+), 5 deletions(-)
Index: 2.6.24-rc3-mm2/drivers/block/floppy.c
This patch converts DAC960 to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/block/DAC960.c |5 +
1 files changed, 1 insertion(+), 4 deletions(-)
Index: 2.6.24-rc3-mm2/drivers/block/DAC960.c
This patch converts um to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
arch/um/drivers/ubd_kern.c | 10 +-
1 files changed, 1 insertion(+), 9 deletions(-)
Index: 2.6.24-rc3-mm2/arch/um/drivers/ubd_kern.c
This patch converts core parts of block layer to use blk_end_request().
'dequeue' argument was originally introduced for end_dequeued_request(),
where no attempt should be made to dequeue the request as it's already
dequeued.
However, it's not necessary as it can be checked with
This patch adds/exports functions to get the size of request in bytes.
They are useful because blk_end_request() takes bytes
as a completed I/O size instead of sectors.
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
block/ll_rw_blk.c |
Richard Scobie wrote:
If one disregards the rotational speed and access time advantage that
SAS drives have over SATA, does the SAS interface offer any performance
advantage?
It's a very good question, to which I wish I have an answer myself ;)
Since I never tried actual SAS controllers with
This patch converts sx8 to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/block/sx8.c |4 +---
1 files changed, 1 insertion(+), 3 deletions(-)
Index: 2.6.24-rc3-mm2/drivers/block/sx8.c
This patch converts arm to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
arch/arm/plat-omap/mailbox.c |9 ++---
1 files changed, 6 insertions(+), 3 deletions(-)
Index:
Basically, yes but I think it would be nice to be able to change the
value on the fly.
If you export it then that will just work. If you want to be able to
set it per device we just need to inherit it at creation time for the
device.
As to the sysfs node for per device setup that makes
Alan Cox wrote:
polling for media change which involves reading maybe a few tens of
bytes. The overhead of doing those using PIO wouldn't be too much more
than the cost of writing CDBs out. It would help more if we can cut
down the number of commands used for testing media changed event (on
Pavel Machek wrote:
On Fri 2007-11-30 13:13:44, Alan Cox wrote:
Why does a single spurious interrupt cause it to be shut down? I can
It doesn't.
see if the interrupt is stuck on and keeps interrupting constantly, but
if it's just the occasional spurious interrupt, why not just ignore it
polling for media change which involves reading maybe a few tens of
bytes. The overhead of doing those using PIO wouldn't be too much more
than the cost of writing CDBs out. It would help more if we can cut
down the number of commands used for testing media changed event (on my
to do list).
On Fri 2007-11-30 13:13:44, Alan Cox wrote:
Why does a single spurious interrupt cause it to be shut down? I can
It doesn't.
see if the interrupt is stuck on and keeps interrupting constantly, but
if it's just the occasional spurious interrupt, why not just ignore it
and move on?
Why does a single spurious interrupt cause it to be shut down? I can
It doesn't.
see if the interrupt is stuck on and keeps interrupting constantly, but
if it's just the occasional spurious interrupt, why not just ignore it
and move on?
The interrupt is usually level triggered so it
One thing I don't understand about this argument is that PIO cycle time
is not determined by CPU power. It's bound by PCI and ATA bus speed.
If you put a faster CPU on the job, it just ends up wasting the same
amount of time burning up more CPU cycles or am I misjudging power of
those
On 11/30/07, Bjoern Olausson [EMAIL PROTECTED] wrote:
On 11/30/07, Tejun Heo [EMAIL PROTECTED] wrote:
Please first verify the problem occurs on your machine before updating
firmware.
Hehe, I am shure it occures... I reported the bug ;-)
( http://bugzilla.kernel.org/show_bug.cgi?id=9010
On 11/30/07, Tejun Heo [EMAIL PROTECTED] wrote:
Please first verify the problem occurs on your machine before updating
firmware.
Hehe, I am shure it occures... I reported the bug ;-)
( http://bugzilla.kernel.org/show_bug.cgi?id=9010 )
You can download needed software from silicon image's web
On Fri, Nov 30, 2007 at 02:05:01PM +0300, Sergei Shtylyov wrote:
Anton Vorontsov wrote:
Remaining question: any preferred name for that property? pio-mode okay?
It's assuming that PIO6 capable bus supports PIO0 as well, thus no mask.
I've already suggested generic. A name simple also
Anton Vorontsov wrote:
Remaining question: any preferred name for that property? pio-mode okay?
It's assuming that PIO6 capable bus supports PIO0 as well, thus no mask.
I've already suggested generic. A name simple also comes to my mind.
You've misread my question. I didn't ask about
Anton Vorontsov wrote:
Remaining question: any preferred name for that property? pio-mode okay?
It's assuming that PIO6 capable bus supports PIO0 as well, thus no mask.
I've already suggested generic. A name simple also comes to my mind.
You've misread my question. I didn't ask about
On 11/30/07, Tejun Heo [EMAIL PROTECTED] wrote:
JMB? That's Jmicron. What should be updated is the SIMG4726 firmware.
Sry you are right seems my mind was on holliday when writing the mail.
Okay, I'll see if I can find something to update the SIMG4726 and test
hotpuging.
(any advices?)
Anton Vorontsov wrote:
Remaining question: any preferred name for that property? pio-mode okay?
It's assuming that PIO6 capable bus supports PIO0 as well, thus no mask.
I've already suggested generic. A name simple also comes to my mind.
WBR, Sergei
-
To unsubscribe from this list: send
Bartlomiej Zolnierkiewicz wrote:
Signed-off-by: Bartlomiej Zolnierkiewicz [EMAIL PROTECTED]
Acked-by: Sergei Shtylyov [EMAIL PROTECTED]
MBR, Sergei
-
To unsubscribe from this list: send the line unsubscribe linux-ide in
the body of a message to [EMAIL PROTECTED]
More majordomo info at
Bartlomiej Zolnierkiewicz wrote:
Signed-off-by: Bartlomiej Zolnierkiewicz [EMAIL PROTECTED]
Acked-by: Sergei Shtylyov [EMAIL PROTECTED]
MBR, Sergei
-
To unsubscribe from this list: send the line unsubscribe linux-ide in
the body of a message to [EMAIL PROTECTED]
More majordomo info at
sam song wrote:
Tejun Heo wrote:
Yeah, that's SLE10 so no wonder ATA passthrough is broken. Please try
with the vanilla kernel.
Still one point to consultant with you. I have no successful experience with
vanilla kernel upgrade. So after getting the kernel 2.6.23.9, is it OK for
me to use
This patch converts mmc to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/mmc/card/block.c | 24 +---
drivers/mmc/card/queue.c |4 ++--
2 files changed, 7 insertions(+), 21 deletions(-)
fr., 30.11.2007 kl. 19.39 +0900, skrev Tejun Heo:
Kjartan Maraas wrote:
on., 28.11.2007 kl. 10.09 +0900, skrev Tejun Heo:
Kjartan Maraas wrote:
I get this exact error message on a normal first time boot here. I'm
using the latest fedora development kernel which is 2.6.24-rc2-git6
Tejun Heo wrote:
Because SFF ATA controller don't have IRQ pending bit. You don't know
whether IRQ is raised or not. Plus, accessing the status register which
clears pending IRQ can be very slow on PATA machines. It has to go
through the PCI and ATA bus and come back. So, unconditionally
On Fri, Nov 30, 2007 at 02:43:50PM +0300, Sergei Shtylyov wrote:
Anton Vorontsov wrote:
Remaining question: any preferred name for that property? pio-mode okay?
It's assuming that PIO6 capable bus supports PIO0 as well, thus no mask.
I've already suggested generic. A name simple also
Tejun Heo wrote:
Yeah, that's SLE10 so no wonder ATA passthrough is broken. Please try
with the vanilla kernel.
Still one point to consultant with you. I have no successful experience with
vanilla kernel upgrade. So after getting the kernel 2.6.23.9, is it OK for
me to use the default
Bjoern Olausson wrote:
On 11/30/07, Tejun Heo [EMAIL PROTECTED] wrote:
JMB? That's Jmicron. What should be updated is the SIMG4726 firmware.
Sry you are right seems my mind was on holliday when writing the mail.
Okay, I'll see if I can find something to update the SIMG4726 and test
Kjartan Maraas wrote:
on., 28.11.2007 kl. 10.09 +0900, skrev Tejun Heo:
Kjartan Maraas wrote:
I get this exact error message on a normal first time boot here. I'm
using the latest fedora development kernel which is 2.6.24-rc2-git6
based. And I have the latest BIOS from HP IIRC.
This is an
Tejun Heo wrote:
Alan Cox wrote:
polling for media change which involves reading maybe a few tens of
bytes. The overhead of doing those using PIO wouldn't be too much more
than the cost of writing CDBs out. It would help more if we can cut
down the number of commands used for testing media
Tejun Heo wrote:
Hello,
Mark Paulus wrote:
sata_sil :02:0b.0: version 2.2
ACPI: PCI Interrupt :02:0b.0[A] - GSI 22 (level, low) - IRQ 18
scsi0 : sata_sil
scsi1 : sata_sil
ata1: SATA max UDMA/100 cmd 0xe0812080 ctl 0xe081208a bmdma 0xe0812000
irq 18
ata2: SATA max UDMA/100 cmd
This patch converts nbd to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/block/nbd.c |4 +---
1 files changed, 1 insertion(+), 3 deletions(-)
Index: 2.6.24-rc3-mm2/drivers/block/nbd.c
This patch converts ub to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/block/ub.c |4 ++--
1 files changed, 2 insertions(+), 2 deletions(-)
Index: 2.6.24-rc3-mm2/drivers/block/ub.c
This patch converts xen-blkfront to use blk_end_request().
Signed-off-by: Kiyoshi Ueda [EMAIL PROTECTED]
Signed-off-by: Jun'ichi Nomura [EMAIL PROTECTED]
---
drivers/block/xen-blkfront.c |5 ++---
1 files changed, 2 insertions(+), 3 deletions(-)
Index:
This patch converts bidi of scsi mid-layer to use blk_end_request().
rq-next_rq represents a pair of bidi requests.
(There are no other use of 'next_rq' of struct request.)
For both requests in the pair, end_that_request_chunk() should be
called before end_that_request_last() is called for one of
The comment I saw, which I'm trying to verify, mentioned the SATA drives
held the bus or similar longer than SAS ones.
SATA normally uses one link per device so the device side isn't contended
unless you descend into the murky world of port multipliers.
On the host side an AHCI controller
Alan Cox wrote:
If you want really high performance use multiple drives, on multiple PCIE
controllers. Just make sure your backup planning of raid 1+0 setup is
done right as many drives means a lot more drive fails.
Thanks again. For what it's worth, I shall be attempting this with SATA
64 matches
Mail list logo
