From: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
To: Lucas De Marchi <lucas.demarchi@intel.com>,
Tvrtko Ursulin <tursulin@igalia.com>
Cc: igt-dev@lists.freedesktop.org,
"Alex Deucher" <alexander.deucher@amd.com>,
"Christian König" <christian.keonig@amd.com>,
"Rob Clark" <robdclark@chromium.org>
Subject: Re: [PATCH i-g-t] lib/igt_drm_fdinfo: Handle amdgpu memory stats
Date: Fri, 3 May 2024 14:53:20 +0100 [thread overview]
Message-ID: <c7272ce2-1932-447f-8455-94f66a313401@igalia.com> (raw)
In-Reply-To: <crxjiae2z6mwas6rxq7cwsekfw26llg5oqlk5motte6qt2kb33@6j6plap5mxu6>
On 03/05/2024 14:32, Lucas De Marchi wrote:
> On Fri, May 03, 2024 at 01:37:31PM GMT, Tvrtko Ursulin wrote:
>> From: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
>>
>> Code so far only handles the clients using the common DRM helper.
>>
>> Handle the amdgpu driver which uses a slightly different set of keys.
>> More
>> specifically, outputs drm-memory-<region> instead of drm-total-<region>.
>>
>> With this added gputop starts showing total memory usage for amdgpu.
>>
>> Signed-off-by: Tvrtko Ursulin <tvrtko.ursulin@igalia.com>
>> Cc: Alex Deucher <alexander.deucher@amd.com>
>> Cc: Christian König <christian.keonig@amd.com>
>> Cc: Rob Clark <robdclark@chromium.org>
>> Cc: Lucas De Marchi <lucas.demarchi@intel.com>
>> ---
>> lib/igt_drm_fdinfo.c | 4 ++++
>> 1 file changed, 4 insertions(+)
>>
>> diff --git a/lib/igt_drm_fdinfo.c b/lib/igt_drm_fdinfo.c
>> index cab677df29a9..89992f1477ec 100644
>> --- a/lib/igt_drm_fdinfo.c
>> +++ b/lib/igt_drm_fdinfo.c
>> @@ -262,6 +262,10 @@ __igt_parse_drm_fdinfo(int dir, const char *fd,
>> struct drm_client_fdinfo *info,
>> idx = parse_region(l, info, strlen("drm-total-"),
>> region_map, region_entries, &val);
>> UPDATE_REGION(idx, total, val);
>> + } else if (!strncmp(l, "drm-memory-", 11)) {
>> + idx = parse_region(l, info, strlen("drm-memory-"),
>> + region_map, region_entries, &val);
>> + UPDATE_REGION(idx, total, val);
>
> are they the same thing? Should we clarify the doc that a driver can't
> use both drm-memory- and drm-total-?
>
> - drm-memory-<region>: <uint> [KiB|MiB]
> Each possible memory type which can be used to store buffer objects
> by the GPU in question shall be given a stable and unique name to be
> returned as the string here. The name "memory" is reserved to refer
> to normal system memory.
> Value shall reflect the amount of storage currently consumed by the
> buffer objects belong to this client, in the respective memory region.
> Default unit shall be bytes with optional unit specifiers of 'KiB'
> or 'MiB' indicating kibi- or mebi-bytes.
> - drm-total-<region>: <uint> [KiB|MiB]
>
> The total size of buffers that including shared and private memory.
>
> If they are the same and the total field can just be re-used,
Yep exactly, they are the same in practice so we are lucky that we are
able to fix it still. See:
https://lore.kernel.org/amd-gfx/CADnq5_NAO_Ao0EJTO=MJxvR-KJkF1WCwKGV-9ami7qQdzf029w@mail.gmail.com/T/#t
(I forgot to CC dri-devel when posting the patch.)
>
>
> Reviewed-by: Lucas De Marchi <lucas.demarchi@intel.com>
Thanks! Are you okay if I put this in first since it collides with your
fdinfo rework?
Although I will wait for the drm-usage-stats.rst patch to get approved
first.
Regards,
Tvrtko
> Lucas De Marchi
>
>> } else if (!strncmp(l, "drm-shared-", 11)) {
>> idx = parse_region(l, info, strlen("drm-shared-"),
>> region_map, region_entries, &val);
>> --
>> 2.44.0
>>
next prev parent reply other threads:[~2024-05-03 13:53 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-05-03 12:37 [PATCH i-g-t] lib/igt_drm_fdinfo: Handle amdgpu memory stats Tvrtko Ursulin
2024-05-03 12:40 ` Tvrtko Ursulin
2024-05-03 13:32 ` Lucas De Marchi
2024-05-03 13:53 ` Tvrtko Ursulin [this message]
2024-05-03 13:59 ` Lucas De Marchi
2024-05-03 15:54 ` ✓ Fi.CI.BAT: success for " Patchwork
2024-05-03 16:29 ` ✓ CI.xeBAT: " Patchwork
2024-05-03 19:14 ` ✗ CI.xeFULL: failure " Patchwork
2024-05-04 3:00 ` ✗ 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=c7272ce2-1932-447f-8455-94f66a313401@igalia.com \
--to=tvrtko.ursulin@igalia.com \
--cc=alexander.deucher@amd.com \
--cc=christian.keonig@amd.com \
--cc=igt-dev@lists.freedesktop.org \
--cc=lucas.demarchi@intel.com \
--cc=robdclark@chromium.org \
--cc=tursulin@igalia.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