Hi Martin,
I will provide an update patch list soon.
Thanks,
~Saurav
-Original Message-
From: "Martin K. Petersen"
Organization: Oracle Corporation
Date: Tuesday, 19 March 2019 at 5:32 AM
To: Saurav Kashyap , Chad Dupuis
Cc: "kbuild-...@01.org" , "martin.peter...@oracle.com"
, "qlogic
On Thu, Mar 21, 2019 at 08:17:44PM -0700, Bart Van Assche wrote:
> On 3/21/19 7:11 PM, Ming Lei wrote:
> > On Fri, Mar 22, 2019 at 09:45:54AM +0800, zhengbin wrote:
> > > + /* If device state changes to SDEV_RUNNING, we need to run hw queue
> > > + * to avoid io hung.
> > > + */
> > > + if (ret =
On 3/21/19 7:11 PM, Ming Lei wrote:
On Fri, Mar 22, 2019 at 09:45:54AM +0800, zhengbin wrote:
+ /* If device state changes to SDEV_RUNNING, we need to run hw queue
+* to avoid io hung.
+*/
+ if (ret == 0 && state == SDEV_RUNNING)
+ blk_mq_run_hw_queues(s
Hi zhengbin,
Thank you for the patch! Perhaps something to improve:
[auto build test WARNING on mkp-scsi/for-next]
[also build test WARNING on v5.1-rc1 next-20190321]
[if your patch is applied to the wrong git tree, please drop us a note to help
improve the system]
url:
https://github.com
On Fri, Mar 22, 2019 at 10:56:46AM +0800, zhengbin wrote:
> When I use dd test a SCSI device which use blk-mq in the following steps:
> 1.echo "blocked" >/sys/block/sda/device/state
> 2.dd if=/dev/sda of=/mnt/t.log bs=1M count=10
> 3.echo "running" >/sys/block/sda/device/state
> dd should finish th
When I use dd test a SCSI device which use blk-mq in the following steps:
1.echo "blocked" >/sys/block/sda/device/state
2.dd if=/dev/sda of=/mnt/t.log bs=1M count=10
3.echo "running" >/sys/block/sda/device/state
dd should finish this work after step 3, unfortunately, still hung.
After step2, the k
On Fri, Mar 22, 2019 at 10:18:54AM +0800, zhengbin (A) wrote:
> So blk_mq_run_hw_queues should still be in mutex_lock(&sdev->state_mutex)??
I think so, this way should be safe given the pattern is used in
__scsi_remove_device() already.
BTW, please do not reply before the quoted text, and I'd sug
So blk_mq_run_hw_queues should still be in mutex_lock(&sdev->state_mutex)??
On 2019/3/22 10:11, Ming Lei wrote:
> On Fri, Mar 22, 2019 at 09:45:54AM +0800, zhengbin wrote:
>> When I use dd test a SCSI device which use blk-mq in the following steps:
>> 1.echo "blocked" >/sys/block/sda/device/state
On Fri, Mar 22, 2019 at 09:45:54AM +0800, zhengbin wrote:
> When I use dd test a SCSI device which use blk-mq in the following steps:
> 1.echo "blocked" >/sys/block/sda/device/state
> 2.dd if=/dev/sda of=/mnt/t.log bs=1M count=10
> 3.echo "running" >/sys/block/sda/device/state
> dd should finish th
On 2019-03-21 7:33 p.m., Bart Van Assche wrote:
On Thu, 2019-03-21 at 19:11 -0400, Douglas Gilbert wrote:
That doesn't sound right. If it was correct then sg_open() and sg_release()
have mutex overkill (and I don't think that is caused by the complexity of
adding O_EXCL which is damn hard to imp
When I use dd test a SCSI device which use blk-mq in the following steps:
1.echo "blocked" >/sys/block/sda/device/state
2.dd if=/dev/sda of=/mnt/t.log bs=1M count=10
3.echo "running" >/sys/block/sda/device/state
dd should finish this work after step 3, unfortunately, still hung.
After step2, the k
I have a business Proposal that will be of benefit to the both of us.Kindly
contact me on mrmichealwu...@yahoo.com.hk should this be of interest to you.
On Thu, 2019-03-21 at 19:11 -0400, Douglas Gilbert wrote:
> That doesn't sound right. If it was correct then sg_open() and sg_release()
> have mutex overkill (and I don't think that is caused by the complexity of
> adding O_EXCL which is damn hard to implement correctly).
>
> Example with existing
On 2019-03-21 7:11 p.m., Douglas Gilbert wrote:
On 2019-03-21 5:36 p.m., Bart Van Assche wrote:
On Wed, 2019-03-20 at 12:39 +0100, Hannes Reinecke wrote:
The 'ch_mutex" is meant to guard against modifications of
file->private_data, so we need to take in in ch_release() as
well as in ch_open().
On 2019-03-21 5:36 p.m., Bart Van Assche wrote:
On Wed, 2019-03-20 at 12:39 +0100, Hannes Reinecke wrote:
The 'ch_mutex" is meant to guard against modifications of
file->private_data, so we need to take in in ch_release() as
well as in ch_open().
Signed-off-by: Hannes Reinecke
---
drivers/sc
On Wed, 2019-03-20 at 20:35 -0400, Martin K. Petersen wrote:
> Bart,
>
> > You are right of course. The reason I would like to inline that
> > function is because it was essential in the past to have that code as
> > a separate function but after patch 1/2 inlining that function makes
> > sd_probe
On Wed, 2019-03-20 at 12:39 +0100, Hannes Reinecke wrote:
> The 'ch_mutex" is meant to guard against modifications of
> file->private_data, so we need to take in in ch_release() as
> well as in ch_open().
>
> Signed-off-by: Hannes Reinecke
> ---
> drivers/scsi/ch.c | 2 ﭗ橸ṷ梧뇪觬()
>
> diff --gi
If the drive has indicated to us that it doesn't have any media in
it, don't bother after the third poll indicating "media_not_present".
Previously we gave up immediately after the first indication, however
two retries were still available for orther cases.
This change solved a problem of mounting
> > > > >>> Hannes & Christoph: Please comment on Sreekanth's proposed
> approach.
> > > > >>
> > > > >> Iterating over all tags from the driver is always wrong. We've
> > > > >> been though this a few times.
> > > > >
> > > > > Current issue is very easy to be reproduced and it is widely
> > > >
On Thu, 2019-03-21 at 13:43 +0800, zhengbin wrote:
> diff --git a/drivers/scsi/scsi_sysfs.c b/drivers/scsi/scsi_sysfs.c
> index 6a9040f..342e513 100644
> --- a/drivers/scsi/scsi_sysfs.c
> +++ b/drivers/scsi/scsi_sysfs.c
> @@ -771,9 +771,20 @@ store_state_field(struct device *dev, struct
> device_a
From: Willy Tarreau
From: Silvio Cesare
Change snprintf to scnprintf. There are generally two cases where using
snprintf causes problems.
1) Uses of size += snprintf(buf, SIZE - size, fmt, ...)
In this case, if snprintf would have written more characters than what the
buffer size (SIZE) is, the
On Thu, 2019-03-21 at 12:20 +0800, jianchao.wang wrote:
> Do you think we should provide more guarantee when setting sdev state through
> sysfs ?
>
> In the current implementation, store_state_field does nothing but just modify
> state.
>
> For example, when we set state to 'blocked', it cannot
On 3/20/19 8:55 PM, Ming Lei wrote:
I suggest to move blk_mq_run_hw_queues() out of the mutex for
avoiding any potential trouble.
Which trouble? I don't think any trouble can arise from holding a mutex
around a request to run a request queue asynchronously.
Bart.
On Wed, 2019-03-20 at 16:35 -0400, Laurence Oberman wrote:
> On Wed, 2019-03-20 at 13:28 -0700, Bart Van Assche wrote:
> > On Wed, 2019-03-20 at 11:11 -0400, Laurence Oberman wrote:
> > > On Wed, 2019-03-20 at 09:45 -0400, Laurence Oberman wrote:
> > > > Hello Bart, I hope all is well with you.
> >
Hi,
we have 4 qemu hosts with identical hardware in a cluster which were
running fine for a year or so. Recently one host started exhibiting
issues with the disks going offline. If it happens all 10 disks
attached to the controller go offline.
Managed to capture some logs, they're a bit lower. Ac
Dear Friend,
I am Ms Safi Kabore work with the department of Audit and accounting manager
here in the Bank,
There is this fund that was keep in my custody years ago,please i need your
assistance for the transferring of thIs fund to your bank account for both of
us benefit for life time inv
On Thu, Mar 21, 2019 at 01:43:09PM +0800, zhengbin wrote:
> When I use dd test a SCSI device which use blk-mq in the following steps:
> 1.echo "blocked" >/sys/block/sda/device/state
> 2.dd if=/dev/sda of=/mnt/t.log bs=1M count=10
> 3.echo "running" >/sys/block/sda/device/state
> dd should finish th
27 matches
Mail list logo
