On 04/20/2018 02:27 AM, James Morse wrote:
> Hi Alex,
>
> On 04/16/2018 10:59 PM, Alex G. wrote:
>> On 04/13/2018 11:38 AM, James Morse wrote:
>>> This assumes a cache-invalidate will clear the error, which I don't
> think we're
>>> guaranteed on arm.
>>> It also destroys any adjacent data,
On 04/20/2018 02:27 AM, James Morse wrote:
> Hi Alex,
>
> On 04/16/2018 10:59 PM, Alex G. wrote:
>> On 04/13/2018 11:38 AM, James Morse wrote:
>>> This assumes a cache-invalidate will clear the error, which I don't
> think we're
>>> guaranteed on arm.
>>> It also destroys any adjacent data,
Hi Alex,
On 04/16/2018 10:59 PM, Alex G. wrote:
> On 04/13/2018 11:38 AM, James Morse wrote:
>> This assumes a cache-invalidate will clear the error, which I don't
think we're
>> guaranteed on arm.
>> It also destroys any adjacent data, "everyone's happy" includes the
thread that
>> got a
Hi Alex,
On 04/16/2018 10:59 PM, Alex G. wrote:
> On 04/13/2018 11:38 AM, James Morse wrote:
>> This assumes a cache-invalidate will clear the error, which I don't
think we're
>> guaranteed on arm.
>> It also destroys any adjacent data, "everyone's happy" includes the
thread that
>> got a
On 04/13/2018 11:38 AM, James Morse wrote:
> Hi Alex,
>
> On 09/04/18 19:11, Alex G. wrote:
>> On 04/06/2018 01:24 PM, James Morse wrote:
>> Do you have any ETA on when your SEA patches are going to make it
>> upstream? There's not much point in updating my patchset if it's going
>> to conflict
On 04/13/2018 11:38 AM, James Morse wrote:
> Hi Alex,
>
> On 09/04/18 19:11, Alex G. wrote:
>> On 04/06/2018 01:24 PM, James Morse wrote:
>> Do you have any ETA on when your SEA patches are going to make it
>> upstream? There's not much point in updating my patchset if it's going
>> to conflict
Hi Alex,
On 09/04/18 19:11, Alex G. wrote:
> On 04/06/2018 01:24 PM, James Morse wrote:
> Do you have any ETA on when your SEA patches are going to make it
> upstream? There's not much point in updating my patchset if it's going
> to conflict with your work.
The SEA stuff went in with
Hi Alex,
On 09/04/18 19:11, Alex G. wrote:
> On 04/06/2018 01:24 PM, James Morse wrote:
> Do you have any ETA on when your SEA patches are going to make it
> upstream? There's not much point in updating my patchset if it's going
> to conflict with your work.
The SEA stuff went in with
Hi James,
On 04/06/2018 01:24 PM, James Morse wrote:
Do you have any ETA on when your SEA patches are going to make it
upstream? There's not much point in updating my patchset if it's going
to conflict with your work.
(snip)
>> On the contrary. No output, followed by a watchdog reboot is
Hi James,
On 04/06/2018 01:24 PM, James Morse wrote:
Do you have any ETA on when your SEA patches are going to make it
upstream? There's not much point in updating my patchset if it's going
to conflict with your work.
(snip)
>> On the contrary. No output, followed by a watchdog reboot is
Hi Alex,
On 04/04/18 20:49, Alex G. wrote:
> On 04/04/2018 11:53 AM, James Morse wrote:
How do we know we will survive this trip?
>>>
>>> We don't.
>>
>> Isn't that even worse than panic()ing? (No output, followed by a watchdog
>> reboot
>> if we're lucky)
> On the contrary. No output,
Hi Alex,
On 04/04/18 20:49, Alex G. wrote:
> On 04/04/2018 11:53 AM, James Morse wrote:
How do we know we will survive this trip?
>>>
>>> We don't.
>>
>> Isn't that even worse than panic()ing? (No output, followed by a watchdog
>> reboot
>> if we're lucky)
> On the contrary. No output,
On 04/04/2018 11:53 AM, James Morse wrote:
> Hi Alex,
(snip)
>>> How do we know we will survive this trip?
>>
>> We don't.
>
> Isn't that even worse than panic()ing? (No output, followed by a watchdog
> reboot
> if we're lucky)
On the contrary. No output, followed by a watchdog reboot is
On 04/04/2018 11:53 AM, James Morse wrote:
> Hi Alex,
(snip)
>>> How do we know we will survive this trip?
>>
>> We don't.
>
> Isn't that even worse than panic()ing? (No output, followed by a watchdog
> reboot
> if we're lucky)
On the contrary. No output, followed by a watchdog reboot is
Hi Alex,
On 04/04/18 16:33, Alex G. wrote:
> On 04/04/2018 02:18 AM, James Morse wrote:
>> On 03/04/18 18:08, Alexandru Gagniuc wrote:
>>> BIOSes like to send NMIs for a number of silly reasons often deemed
>>> to be "fatal". For example pin bounce during a PCIE hotplug/unplug
>>> might cause the
Hi Alex,
On 04/04/18 16:33, Alex G. wrote:
> On 04/04/2018 02:18 AM, James Morse wrote:
>> On 03/04/18 18:08, Alexandru Gagniuc wrote:
>>> BIOSes like to send NMIs for a number of silly reasons often deemed
>>> to be "fatal". For example pin bounce during a PCIE hotplug/unplug
>>> might cause the
Hi James,
On 04/04/2018 02:18 AM, James Morse wrote:
> Hi Alexandru,
>
> On 03/04/18 18:08, Alexandru Gagniuc wrote:
>> BIOSes like to send NMIs for a number of silly reasons often deemed
>> to be "fatal". For example pin bounce during a PCIE hotplug/unplug
>> might cause the link to go down and
Hi James,
On 04/04/2018 02:18 AM, James Morse wrote:
> Hi Alexandru,
>
> On 03/04/18 18:08, Alexandru Gagniuc wrote:
>> BIOSes like to send NMIs for a number of silly reasons often deemed
>> to be "fatal". For example pin bounce during a PCIE hotplug/unplug
>> might cause the link to go down and
Hi Alexandru,
On 03/04/18 18:08, Alexandru Gagniuc wrote:
> BIOSes like to send NMIs for a number of silly reasons often deemed
> to be "fatal". For example pin bounce during a PCIE hotplug/unplug
> might cause the link to go down and retrain, with fatal PCI errors
> being generated while the
Hi Alexandru,
On 03/04/18 18:08, Alexandru Gagniuc wrote:
> BIOSes like to send NMIs for a number of silly reasons often deemed
> to be "fatal". For example pin bounce during a PCIE hotplug/unplug
> might cause the link to go down and retrain, with fatal PCI errors
> being generated while the
BIOSes like to send NMIs for a number of silly reasons often deemed
to be "fatal". For example pin bounce during a PCIE hotplug/unplug
might cause the link to go down and retrain, with fatal PCI errors
being generated while the link is retraining.
Instead of panic()ing in NMI context, pass fatal
BIOSes like to send NMIs for a number of silly reasons often deemed
to be "fatal". For example pin bounce during a PCIE hotplug/unplug
might cause the link to go down and retrain, with fatal PCI errors
being generated while the link is retraining.
Instead of panic()ing in NMI context, pass fatal
22 matches
Mail list logo