[PATCH 0/3] scsi: arcmsr: add driver module parameter - msi_enable, msix_enable

[PATCH 0/3] scsi: arcmsr: add driver module parameter - msi_enable, msix_enable

[Ching Huang]

From: Ching Huang <ching2048@areca.com.tw>

Hi all,

The following patches apply to Martin's 4.16/scsi-queue.

Patch 1: Add module parameter msi_enable to has a chance to disable msi interrupt if it does not work properly.

Patch 2: Add module parameter msix_enable to has a chance to disable msix interrupt if it does not work properly.

Patch 3: Update driver version to v1.40.00.03-20171121

Please review. Thanks.
---

Re: [PATCH 0/3] scsi: arcmsr: add driver module parameter - msi_enable, msix_enable

[Christoph Hellwig]

On Thu, Nov 23, 2017 at 09:22:03AM +0800, Ching Huang wrote:
> From: Ching Huang <ching2048@areca.com.tw>
> 
> Hi all,
> 
> The following patches apply to Martin's 4.16/scsi-queue.
> 
> Patch 1: Add module parameter msi_enable to has a chance to disable msi interrupt if it does not work properly.
> 
> Patch 2: Add module parameter msix_enable to has a chance to disable msix interrupt if it does not work properly.

Why would it not work properly?

Re: [PATCH 0/3] scsi: arcmsr: add driver module parameter - msi_enable, msix_enable

[Ching Huang]

On Thu, 2017-11-23 at 04:57 -0800, Christoph Hellwig wrote:
> On Thu, Nov 23, 2017 at 09:22:03AM +0800, Ching Huang wrote:
> > From: Ching Huang <ching2048@areca.com.tw>
> > 
> > Hi all,
> > 
> > The following patches apply to Martin's 4.16/scsi-queue.
> > 
> > Patch 1: Add module parameter msi_enable to has a chance to disable msi interrupt if it does not work properly.
> > 
> > Patch 2: Add module parameter msix_enable to has a chance to disable msix interrupt if it does not work properly.
> 
> Why would it not work properly?
This patch is apply to
https://git.kernel.org/pub/scm/linux/kernel/git/mkp/scsi.git/tree/?h=4.16/scsi-queue

Re: [PATCH 0/3] scsi: arcmsr: add driver module parameter - msi_enable, msix_enable

[Dan Carpenter]

On Fri, Nov 24, 2017 at 03:12:39AM +0800, Ching Huang wrote:
> On Thu, 2017-11-23 at 04:57 -0800, Christoph Hellwig wrote:
> > On Thu, Nov 23, 2017 at 09:22:03AM +0800, Ching Huang wrote:
> > > From: Ching Huang <ching2048@areca.com.tw>
> > > 
> > > Hi all,
> > > 
> > > The following patches apply to Martin's 4.16/scsi-queue.
> > > 
> > > Patch 1: Add module parameter msi_enable to has a chance to disable msi interrupt if it does not work properly.
> > > 
> > > Patch 2: Add module parameter msix_enable to has a chance to disable msix interrupt if it does not work properly.
> > 
> > Why would it not work properly?
> This patch is apply to
> https://git.kernel.org/pub/scm/linux/kernel/git/mkp/scsi.git/tree/?h=4.16/scsi-queue
>

No, the question is we're adding an option called "msi_enable" which is
used to disable MSI interrupts "if it does not work properly".  Why is
the current code not working properly?

Is there a crash or a performance issue?  What does the bug in the
current code look like from a user perspective?  Can you send us a dmesg
from a failing system?

regards,
dan carpenter

Re: [PATCH 0/3] scsi: arcmsr: add driver module parameter - (linux-kernel: dan.carpenter@oracle.com exclusive) msi_enable, msix_enable

[Kristian Rasmussen]

On 24.11.2017 15:53, Dan Carpenter - dan.carpenter@oracle.com wrote:
> Is there a crash or a performance issue?  What does the bug in the
> current code look like from a user perspective?
It looks like this: https://bugzilla.kernel.org/show_bug.cgi?id=197877

dmesg output can be found in the bug report.

Disabling MSI interrupts with msi_enable=0 (and Ching's patches) does
resolve the issue, but whether it should be considered a fix or a
workaround is not for me to say.

Regards,

Kristian Rasmussen

Re: [PATCH 0/3] scsi: arcmsr: add driver module parameter - msi_enable, msix_enable

[黃清隆]

Dan,

There is our user feedback said that they have some system have MSI
interrupt issue.(but some system are working)
Those have MSI issue system all are old system, even if MSI interrupt
enable and request are successful, there is not any hardware interrupt
for scsi command. Cause each scsi command timeout. So msi_enable
parameter is a way to disable msi interrupt request.
Using legacy interrupt, those have MSI interrupt issue system are
working properly.

Thanks,
Ching

2017-11-24 22:53 GMT+08:00 Dan Carpenter <dan.carpenter@oracle.com>:
> On Fri, Nov 24, 2017 at 03:12:39AM +0800, Ching Huang wrote:
>> On Thu, 2017-11-23 at 04:57 -0800, Christoph Hellwig wrote:
>> > On Thu, Nov 23, 2017 at 09:22:03AM +0800, Ching Huang wrote:
>> > > From: Ching Huang <ching2048@areca.com.tw>
>> > >
>> > > Hi all,
>> > >
>> > > The following patches apply to Martin's 4.16/scsi-queue.
>> > >
>> > > Patch 1: Add module parameter msi_enable to has a chance to disable msi interrupt if it does not work properly.
>> > >
>> > > Patch 2: Add module parameter msix_enable to has a chance to disable msix interrupt if it does not work properly.
>> >
>> > Why would it not work properly?
>> This patch is apply to
>> https://git.kernel.org/pub/scm/linux/kernel/git/mkp/scsi.git/tree/?h=4.16/scsi-queue
>>
>
> No, the question is we're adding an option called "msi_enable" which is
> used to disable MSI interrupts "if it does not work properly".  Why is
> the current code not working properly?
>
> Is there a crash or a performance issue?  What does the bug in the
> current code look like from a user perspective?  Can you send us a dmesg
> from a failing system?
>
> regards,
> dan carpenter