public inbox for intel-gfx@lists.freedesktop.org
 help / color / mirror / Atom feed
From: "Das, Nirmoy" <nirmoy.das@intel.com>
To: Andrzej Hajda <andrzej.hajda@intel.com>,
	"Das, Nirmoy" <nirmoy.das@linux.intel.com>,
	<intel-gfx@lists.freedesktop.org>
Cc: chris.p.wilson@linux.intel.com,
	Matthew Auld <matthew.auld@intel.com>,
	Rodrigo Vivi <rodrigo.vivi@intel.com>
Subject: Re: [Intel-gfx] [PATCH v2] drm/i915/gt: prevent forcewake releases during BAR resize
Date: Thu, 9 Mar 2023 09:31:22 +0100	[thread overview]
Message-ID: <e9638e96-a1ad-e5c1-5141-c1457b08326e@intel.com> (raw)
In-Reply-To: <baa52f92-3a17-ce60-9ed1-39e12b396f13@intel.com>


On 3/9/2023 8:54 AM, Andrzej Hajda wrote:
> On 08.03.2023 18:29, Das, Nirmoy wrote:
>>
>> On 3/8/2023 2:36 PM, Andrzej Hajda wrote:
>>> Tests on DG2 machines show that releasing forcewakes during BAR resize
>>> results later in forcewake ack timeouts.
>> Do we have a fdo/issues url for that? Having that as References would 
>> be nice.
>
> The patch is result of issue reported internally, but...grepping 
> bugtracker I've found potential candidates:
>
> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/6530
> Closes: https://gitlab.freedesktop.org/drm/intel/-/issues/7853
>
> Maybe it would be good to add them on merge.

Works for me.


Thanks,

Nirmoy

> Anyway thx for a-b.
>
> Regards
> Andrzej
>
>>>   Since forcewakes can be realeased
>>> asynchronously the simplest way to prevent it is to get all forcewakes
>>> for duration of BAR resizing.
>>>
>>> v2: hold rpm as well during resizing (Rodrigo)
>>>
>>> Signed-off-by: Andrzej Hajda <andrzej.hajda@intel.com>
>>
>>
>> Acked-by: Nirmoy Das <nirmoy.das@intel.com>
>>
>>
>>> ---
>>> Please ignore resend of v1, my mistake.
>>>
>>> Regards
>>> Andrzej
>>> ---
>>>   drivers/gpu/drm/i915/gt/intel_region_lmem.c | 25 
>>> +++++++++++++++------
>>>   1 file changed, 18 insertions(+), 7 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/i915/gt/intel_region_lmem.c 
>>> b/drivers/gpu/drm/i915/gt/intel_region_lmem.c
>>> index 89fdfc67f8d1e0..2a3217e2890fc7 100644
>>> --- a/drivers/gpu/drm/i915/gt/intel_region_lmem.c
>>> +++ b/drivers/gpu/drm/i915/gt/intel_region_lmem.c
>>> @@ -54,6 +54,7 @@ static void i915_resize_lmem_bar(struct 
>>> drm_i915_private *i915, resource_size_t
>>>       struct resource *root_res;
>>>       resource_size_t rebar_size;
>>>       resource_size_t current_size;
>>> +    intel_wakeref_t wakeref;
>>>       u32 pci_cmd;
>>>       int i;
>>> @@ -102,15 +103,25 @@ static void i915_resize_lmem_bar(struct 
>>> drm_i915_private *i915, resource_size_t
>>>           return;
>>>       }
>>> -    /* First disable PCI memory decoding references */
>>> -    pci_read_config_dword(pdev, PCI_COMMAND, &pci_cmd);
>>> -    pci_write_config_dword(pdev, PCI_COMMAND,
>>> -                   pci_cmd & ~PCI_COMMAND_MEMORY);
>>> +    /*
>>> +     * Releasing forcewake during BAR resizing results in later 
>>> forcewake
>>> +     * ack timeouts and former can happen any time - it is 
>>> asynchronous.
>>> +     * Grabbing all forcewakes prevents it.
>>> +     */
>>> +    with_intel_runtime_pm(i915->uncore.rpm, wakeref) {
>>> +        intel_uncore_forcewake_get(&i915->uncore, FORCEWAKE_ALL);
>>> -    _resize_bar(i915, GEN12_LMEM_BAR, rebar_size);
>>> +        /* First disable PCI memory decoding references */
>>> +        pci_read_config_dword(pdev, PCI_COMMAND, &pci_cmd);
>>> +        pci_write_config_dword(pdev, PCI_COMMAND,
>>> +                       pci_cmd & ~PCI_COMMAND_MEMORY);
>>> -    pci_assign_unassigned_bus_resources(pdev->bus);
>>> -    pci_write_config_dword(pdev, PCI_COMMAND, pci_cmd);
>>> +        _resize_bar(i915, GEN12_LMEM_BAR, rebar_size);
>>> +
>>> +        pci_assign_unassigned_bus_resources(pdev->bus);
>>> +        pci_write_config_dword(pdev, PCI_COMMAND, pci_cmd);
>>> +        intel_uncore_forcewake_put(&i915->uncore, FORCEWAKE_ALL);
>>> +    }
>>>   }
>>>   #else
>>>   static void i915_resize_lmem_bar(struct drm_i915_private *i915, 
>>> resource_size_t lmem_size) {}
>

  reply	other threads:[~2023-03-09  8:31 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-03-08 11:23 [Intel-gfx] [PATCH] drm/i915/gt: prevent forcewake releases during BAR resize Andrzej Hajda
2023-03-08 13:16 ` Rodrigo Vivi
2023-03-08 13:34   ` Andrzej Hajda
2023-03-08 13:36   ` [Intel-gfx] [PATCH v2] " Andrzej Hajda
2023-03-08 17:29     ` Das, Nirmoy
2023-03-09  7:54       ` Andrzej Hajda
2023-03-09  8:31         ` Das, Nirmoy [this message]
2023-03-08 17:31     ` Rodrigo Vivi
2023-03-09  0:11     ` Andi Shyti
2023-03-16 18:20     ` Andrzej Hajda
2023-03-08 15:44 ` [Intel-gfx] ✗ Fi.CI.BAT: failure for drm/i915/gt: prevent forcewake releases during BAR resize (rev3) Patchwork
2023-03-08 16:08   ` Andrzej Hajda
2023-03-09 20:15 ` [Intel-gfx] ✓ Fi.CI.BAT: success for drm/i915/gt: prevent forcewake releases during BAR resize (rev5) Patchwork
2023-03-11 19:06 ` [Intel-gfx] ✓ Fi.CI.IGT: " Patchwork

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=e9638e96-a1ad-e5c1-5141-c1457b08326e@intel.com \
    --to=nirmoy.das@intel.com \
    --cc=andrzej.hajda@intel.com \
    --cc=chris.p.wilson@linux.intel.com \
    --cc=intel-gfx@lists.freedesktop.org \
    --cc=matthew.auld@intel.com \
    --cc=nirmoy.das@linux.intel.com \
    --cc=rodrigo.vivi@intel.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