linux-scsi.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Tomas Henzl <thenzl@redhat.com>
To: Sumit Saxena <sumit.saxena@avagotech.com>, linux-scsi@vger.kernel.org
Cc: Kashyap Desai <kashyap.desai@avagotech.com>
Subject: Re: [PATCH] regression, megaraid - fix irq setup process
Date: Tue, 02 Jun 2015 11:50:27 +0200	[thread overview]
Message-ID: <556D7C63.5020904@redhat.com> (raw)
In-Reply-To: <3c1a3409bc0508d38d3ded5a2d4ec330@mail.gmail.com>

On 06/02/2015 09:06 AM, Sumit Saxena wrote:
>> -----Original Message-----
>> From: Tomas Henzl [mailto:thenzl@redhat.com]
>> Sent: Monday, June 01, 2015 10:57 PM
>> To: linux-scsi@vger.kernel.org
>> Cc: kashyap.desai@avagotech.com; Sumit.Saxena@avagotech.com
>> Subject: [PATCH] regression, megaraid - fix irq setup process
>>
>> This fixes a regression caused by commit
>> d3557fc8be11d25f316884581f487684f8e7dad3
>> megaraid_sas : Add separate function for setting up IRQs This makes boot
> end
>> with 'root does not exist' message on certain adapters.
>>
>> The bug is that the driver does not setup ints for cards without msi-x
> support.
>> This patch fixes it, in addition to that it moves irq-enable call after
> tasklet
>> initialisation - otherwise a kernel panic may occur, when an int arrives
> before
>> the tasklet is ready.
> 
> Thanks for pointing it out.
> Tasklet is only scheduled for MFI adapters inside ISR routine and all
> commands fired prior to "tasklet_init" are in polled mode  and MFI
> adapters will not  send interrupt for the same.
> Fusion adapters may send the interrupts but their ISR function don't use
> tasklet at all so there should not be any problem.
On my card it happens - and kernel panics.

> We will move tasklet_init prior to enable_intr.  Calling enable_intr later
> after firing few more DCMDs will cause more DCMDs to be fired in polled
> mode(though fusion adapter will send interrupt but will not free up MFI
> frame used).
> I agree that cleaning up of this is required. I will send a patch to clean
> it up and address few additional stuff.
This patch fixes a regression caused by your previous patch, a solution
is needed soon.
Please don't forget that this patch fixes also the issue of
the driver not setting the irq at all for cards without msi-x support
and add it to you fix.

--tm
> 
>>
>> Signed-off-by: Tomas Henzl <thenzl@redhat.com>
>> ---
>> drivers/scsi/megaraid/megaraid_sas_base.c | 22 ++++++++++------------
>> 1 file changed, 10 insertions(+), 12 deletions(-)
>>
>> diff --git a/drivers/scsi/megaraid/megaraid_sas_base.c
>> b/drivers/scsi/megaraid/megaraid_sas_base.c
>> index a022c39153..3771d6fff5 100644
>> --- a/drivers/scsi/megaraid/megaraid_sas_base.c
>> +++ b/drivers/scsi/megaraid/megaraid_sas_base.c
>> @@ -4619,18 +4619,16 @@ static int megasas_init_fw(struct
>> megasas_instance *instance)
>> 			instance->msix_vectors = i;
>> 		else
>> 			instance->msix_vectors = 0;
>> +	}
>>
>> -		dev_info(&instance->pdev->dev,
>> -			"firmware supports msix\t: (%d)", fw_msix_count);
>> -		dev_info(&instance->pdev->dev,
>> -			"current msix/online cpus\t: (%d/%d)\n",
>> -			instance->msix_vectors, (unsigned
>> int)num_online_cpus());
>> +	dev_info(&instance->pdev->dev, "firmware supports msix\t: (%d)",
>> +		fw_msix_count);
>> +	dev_info(&instance->pdev->dev, "current msix/online cpus\t:
>> (%d/%d)\n",
>> +		instance->msix_vectors, (unsigned int)num_online_cpus());
>>
>> -		if (instance->msix_vectors ?
>> -				megasas_setup_irqs_msix(instance, 1) :
>> -				megasas_setup_irqs_ioapic(instance))
>> -			goto fail_setup_irqs;
>> -	}
>> +	if (instance->msix_vectors ? megasas_setup_irqs_msix(instance, 1)
> :
>> +				     megasas_setup_irqs_ioapic(instance))
>> +		goto fail_setup_irqs;
>>
>> 	instance->ctrl_info = kzalloc(sizeof(struct megasas_ctrl_info),
>> 				GFP_KERNEL);
>> @@ -4646,8 +4644,6 @@ static int megasas_init_fw(struct megasas_instance
>> *instance)
>> 	/* Get operational params, sge flags, send init cmd to controller
> */
>> 	if (instance->instancet->init_adapter(instance))
>> 		goto fail_init_adapter;
>> -	instance->instancet->enable_intr(instance);
>> -
>> 	printk(KERN_ERR "megasas: INIT adapter done\n");
>>
>> 	/** for passthrough
>> @@ -4769,6 +4765,8 @@ static int megasas_init_fw(struct megasas_instance
>> *instance)
>> 	tasklet_init(&instance->isr_tasklet, instance->instancet->tasklet,
>> 		(unsigned long)instance);
>>
>> +	instance->instancet->enable_intr(instance);
>> +
>> 	/* Launch SR-IOV heartbeat timer */
>> 	if (instance->requestorId) {
>> 		if (!megasas_sriov_start_heartbeat(instance, 1))
> 
>> --
>> 1.9.3
> --
> To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 


  reply	other threads:[~2015-06-02  9:50 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-06-01 17:26 [PATCH] regression, megaraid - fix irq setup process Tomas Henzl
2015-06-02  7:06 ` Sumit Saxena
2015-06-02  9:50   ` Tomas Henzl [this message]
2015-06-02 10:56     ` Sumit Saxena

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=556D7C63.5020904@redhat.com \
    --to=thenzl@redhat.com \
    --cc=kashyap.desai@avagotech.com \
    --cc=linux-scsi@vger.kernel.org \
    --cc=sumit.saxena@avagotech.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).