[PATCH RFC] igb: Fix false positive return of igb_get_auto_rd_done for 82580

[PATCH RFC] igb: Fix false positive return of igb_get_auto_rd_done for 82580

[Guenter Roeck]

From: Tong Ho <tong.ho@ericsson.com>

82580 re-reads the port specific portion of eeprom after port reset.
If called immediately after a reset, igb_get_auto_rd_done() returns
false positive because the done bit has yet to transition from 1 to 0.

Add wrfl() immediately after resetting 82580 port or device,
plus a 1ms delay, to avoid the problem.

Signed-off-by: Tong Ho <tong.ho@ericsson.com>
Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
---
Sent as RFC since I am not entirely sure if the solution is the correct one
to address the problem we are seeing. If there is a better solution, please
let me know.

 drivers/net/igb/e1000_82575.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/igb/e1000_82575.c b/drivers/net/igb/e1000_82575.c
index 0f563c8..28dae48 100644
--- a/drivers/net/igb/e1000_82575.c
+++ b/drivers/net/igb/e1000_82575.c
@@ -1735,10 +1735,10 @@ static s32 igb_reset_hw_82580(struct e1000_hw *hw)
 		ctrl |= E1000_CTRL_RST;
 
 	wr32(E1000_CTRL, ctrl);
+	wrfl();
 
 	/* Add delay to insure DEV_RST has time to complete */
-	if (global_device_reset)
-		msleep(5);
+	msleep(global_device_reset ? 5 : 1);
 
 	ret_val = igb_get_auto_rd_done(hw);
 	if (ret_val) {
-- 
1.7.3.1


------------------------------------------------------------------------------
EditLive Enterprise is the world's most technically advanced content
authoring tool. Experience the power of Track Changes, Inline Image
Editing and ensure content is compliant with Accessibility Checking.
http://p.sf.net/sfu/ephox-dev2dev
_______________________________________________
E1000-devel mailing list
E1000-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel&#174; Ethernet, visit http://communities.intel.com/community/wired

Re: [PATCH RFC] igb: Fix false positive return of igb_get_auto_rd_done for 82580

[Jeff Kirsher]

[-- Attachment #1: Type: text/plain, Size: 877 bytes --]

On Tue, 2011-06-21 at 12:02 -0700, Guenter Roeck wrote:
> From: Tong Ho <tong.ho@ericsson.com>
> 
> 82580 re-reads the port specific portion of eeprom after port reset.
> If called immediately after a reset, igb_get_auto_rd_done() returns
> false positive because the done bit has yet to transition from 1 to 0.
> 
> Add wrfl() immediately after resetting 82580 port or device,
> plus a 1ms delay, to avoid the problem.
> 
> Signed-off-by: Tong Ho <tong.ho@ericsson.com>
> Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
> ---
> Sent as RFC since I am not entirely sure if the solution is the
> correct one
> to address the problem we are seeing. If there is a better solution,
> please
> let me know. 

Thank you for the suggested patch.  Carolyn is the maintainer for igb
and we will look into this issue you are seeing and the suggested fix. 

[-- Attachment #2: This is a digitally signed message part --]
[-- Type: application/pgp-signature, Size: 490 bytes --]

Re: [PATCH RFC] igb: Fix false positive return of igb_get_auto_rd_done for 82580

[Guenter Roeck]

On Tue, Jun 21, 2011 at 05:17:26PM -0400, Jeff Kirsher wrote:
> On Tue, 2011-06-21 at 12:02 -0700, Guenter Roeck wrote:
> > From: Tong Ho <tong.ho@ericsson.com>
> > 
> > 82580 re-reads the port specific portion of eeprom after port reset.
> > If called immediately after a reset, igb_get_auto_rd_done() returns
> > false positive because the done bit has yet to transition from 1 to 0.
> > 
> > Add wrfl() immediately after resetting 82580 port or device,
> > plus a 1ms delay, to avoid the problem.
> > 
> > Signed-off-by: Tong Ho <tong.ho@ericsson.com>
> > Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
> > ---
> > Sent as RFC since I am not entirely sure if the solution is the
> > correct one
> > to address the problem we are seeing. If there is a better solution,
> > please
> > let me know. 
> 
> Thank you for the suggested patch.  Carolyn is the maintainer for igb
> and we will look into this issue you are seeing and the suggested fix. 

Hi Jeff and Carolyn,

Any update ?

Thanks,
Guenter

------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security 
threats, fraudulent activity, and more. Splunk takes this data and makes 
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
_______________________________________________
E1000-devel mailing list
E1000-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/e1000-devel
To learn more about Intel&#174; Ethernet, visit http://communities.intel.com/community/wired

RE: [PATCH RFC] igb: Fix false positive return of igb_get_auto_rd_done for 82580

[Wyborny, Carolyn]

>-----Original Message-----
>From: Guenter Roeck [mailto:guenter.roeck@ericsson.com]
>Sent: Tuesday, July 05, 2011 7:31 AM
>To: Kirsher, Jeffrey T
>Cc: Brandeburg, Jesse; Wyborny, Carolyn; Duyck, Alexander H; Ronciak,
>John; e1000-devel@lists.sourceforge.net; netdev@vger.kernel.org; Tong Ho
>Subject: Re: [PATCH RFC] igb: Fix false positive return of
>igb_get_auto_rd_done for 82580
>
>On Tue, Jun 21, 2011 at 05:17:26PM -0400, Jeff Kirsher wrote:
>> On Tue, 2011-06-21 at 12:02 -0700, Guenter Roeck wrote:
>> > From: Tong Ho <tong.ho@ericsson.com>
>> >
>> > 82580 re-reads the port specific portion of eeprom after port reset.
>> > If called immediately after a reset, igb_get_auto_rd_done() returns
>> > false positive because the done bit has yet to transition from 1 to
>0.
>> >
>> > Add wrfl() immediately after resetting 82580 port or device,
>> > plus a 1ms delay, to avoid the problem.
>> >
>> > Signed-off-by: Tong Ho <tong.ho@ericsson.com>
>> > Signed-off-by: Guenter Roeck <guenter.roeck@ericsson.com>
>> > ---
>> > Sent as RFC since I am not entirely sure if the solution is the
>> > correct one
>> > to address the problem we are seeing. If there is a better solution,
>> > please
>> > let me know.
>>
>> Thank you for the suggested patch.  Carolyn is the maintainer for igb
>> and we will look into this issue you are seeing and the suggested fix.
>
>Hi Jeff and Carolyn,
>
>Any update ?
>
>Thanks,
>Guenter

Sorry for the delay, we've had some holidays, its just passed internal testing and should be coming upstream shortly.

Thanks,

Carolyn

Carolyn Wyborny
Linux Development
LAN Access Division
Intel Corporation