Igt-dev Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Niranjana Vishwanathapura <niranjana.vishwanathapura@intel.com>
To: "Kamil Konieczny" <kamil.konieczny@linux.intel.com>,
	igt-dev@lists.freedesktop.org,
	"Zbigniew Kempczyński" <zbigniew.kempczynski@intel.com>
Subject: Re: [igt-dev] [PATCH i-g-t v4 2/3] lib/intel_compute: Update intel_compute to run on specified engine
Date: Thu, 7 Dec 2023 11:25:17 -0800	[thread overview]
Message-ID: <ZXIcHd6Nl6aSHiPq@nvishwa1-DESK> (raw)
In-Reply-To: <20231207115618.xojuxvgphjqe2sme@kamilkon-desk.igk.intel.com>

On Thu, Dec 07, 2023 at 12:56:18PM +0100, Kamil Konieczny wrote:
>Hi Niranjana,
>On 2023-12-06 at 12:00:54 -0800, Niranjana Vishwanathapura wrote:
>> With CCS_MODE setting, available compute slices can be assigned
>> to specific compute engines. Update intel_compute library to
>> be able to run compute kernel on specified compute or render
>> engine.
>>
>> Signed-off-by: Niranjana Vishwanathapura <niranjana.vishwanathapura@intel.com>
>> Reviewed-by: Zbigniew Kempczyński <zbigniew.kempczynski@intel.com>

<snip>

>> --- a/lib/intel_compute.h
>> +++ b/lib/intel_compute.h
>> @@ -9,6 +9,8 @@
>>  #ifndef INTEL_COMPUTE_H
>>  #define INTEL_COMPUTE_H
>>
>> +#include "xe_drm.h"
>
>With this it looks like Xe only header.
>

Hmm...
Not sure if we can keep these library headers purly driver (i915/xe) agnostic
for long. The library *.c files already include xe_drm.h and deal xe drivers
differently. I am seeing some library header files include structures
defined in xe_drm.h, but let the *.c files to include xe_drm.h header (cheeky
in my opinion).

>> +
>>  /*
>>   * OpenCL Kernels are generated using:
>>   *
>> @@ -28,5 +30,6 @@ struct intel_compute_kernels {
>>  extern const struct intel_compute_kernels intel_compute_square_kernels[];
>>
>>  bool run_intel_compute_kernel(int fd);
>> +bool run_intel_compute_kernel_on_engine(int fd, struct drm_xe_engine_class_instance *eci);
>
>What about using *void here?
>

Not sure if masking it as 'void *' is a good approach.
Besides, I don't think we are going to support this for i915 (not worth the
effort, validation etc). And I don't think we are going to compile the igt
repo without xe_drm.h.
I can add a function descriptor saying that this function is only supported
for xe driver and return error if this function is ever invoked with non-xe
driver.

Is that fine?

>>
>>  #endif	/* INTEL_COMPUTE_H */
>> --
>> 2.21.0.rc0.32.g243a4c7e27
>--  ^^
>Please update your git to at least 2.3x or 2.4x (current lastest is 2.43.0).

Sure.

Thanks,
Niranjana

>
>Regards,
>Kamil

  reply	other threads:[~2023-12-07 19:25 UTC|newest]

Thread overview: 13+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-12-06 20:00 [igt-dev] [PATCH i-g-t v4 0/3] xe: Validate fixed CCS mode setting Niranjana Vishwanathapura
2023-12-06 20:00 ` [igt-dev] [PATCH i-g-t v4 1/3] lib/xe: Add __xe_exec_queue_create() Niranjana Vishwanathapura
2023-12-07 11:41   ` Kamil Konieczny
2023-12-07 19:16     ` Niranjana Vishwanathapura
2023-12-06 20:00 ` [igt-dev] [PATCH i-g-t v4 2/3] lib/intel_compute: Update intel_compute to run on specified engine Niranjana Vishwanathapura
2023-12-07 11:56   ` Kamil Konieczny
2023-12-07 19:25     ` Niranjana Vishwanathapura [this message]
2023-12-06 20:00 ` [igt-dev] [PATCH i-g-t v4 3/3] tests/xe_compute: Validate ccs_mode setting Niranjana Vishwanathapura
2023-12-06 21:53 ` [igt-dev] ✗ Fi.CI.BAT: failure for xe: Validate fixed CCS mode setting Patchwork
2023-12-06 23:28 ` [igt-dev] ✓ CI.xeBAT: success " Patchwork
2023-12-06 23:39 ` [igt-dev] ✓ Fi.CI.BAT: success for xe: Validate fixed CCS mode setting (rev2) Patchwork
2023-12-07  1:13 ` [igt-dev] ✗ Fi.CI.IGT: failure " Patchwork
2023-12-07  1:28 ` [igt-dev] ✓ CI.xeBAT: success " 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=ZXIcHd6Nl6aSHiPq@nvishwa1-DESK \
    --to=niranjana.vishwanathapura@intel.com \
    --cc=igt-dev@lists.freedesktop.org \
    --cc=kamil.konieczny@linux.intel.com \
    --cc=zbigniew.kempczynski@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