From: Michal Wajdeczko <michal.wajdeczko@intel.com>
To: Ilia Levi <ilia.levi@intel.com>, intel-xe@lists.freedesktop.org
Cc: jonathan.cavitt@intel.com, niranjana.vishwanathapura@intel.com,
matthew.brost@intel.com, koby.elbaz@intel.com,
yaron.avizrat@intel.com
Subject: Re: [PATCH v2 5/5] drm/xe: separate memirq reset/postinstall from vf
Date: Wed, 28 Aug 2024 22:42:32 +0200 [thread overview]
Message-ID: <1429c1eb-91b8-43e0-9e3d-51b2367f2b9d@intel.com> (raw)
In-Reply-To: <20240828091841.1840086-6-ilia.levi@intel.com>
On 28.08.2024 11:18, Ilia Levi wrote:
> Up until now only VF used Memory Based Interrupts (memirq).
> Moving reset/postinstall flows out of VF to cater for other usages.
>
> Signed-off-by: Ilia Levi <ilia.levi@intel.com>
> Reviewed-by: Jonathan Cavitt <jonathan.cavitt@intel.com>
> ---
> drivers/gpu/drm/xe/xe_device.c | 2 +-
> drivers/gpu/drm/xe/xe_device.h | 6 ++++++
> drivers/gpu/drm/xe/xe_irq.c | 36 +++++++++++++++++++---------------
> 3 files changed, 27 insertions(+), 17 deletions(-)
>
> diff --git a/drivers/gpu/drm/xe/xe_device.c b/drivers/gpu/drm/xe/xe_device.c
> index c4f8bfd08b09..a23fb0c81ca6 100644
> --- a/drivers/gpu/drm/xe/xe_device.c
> +++ b/drivers/gpu/drm/xe/xe_device.c
> @@ -660,7 +660,7 @@ int xe_device_probe(struct xe_device *xe)
> err = xe_ggtt_init_early(tile->mem.ggtt);
> if (err)
> return err;
> - if (IS_SRIOV_VF(xe)) {
> + if (xe_device_needs_memirq(xe)) {
> err = xe_memirq_init(&tile->memirq, false);
> if (err)
> return err;
> diff --git a/drivers/gpu/drm/xe/xe_device.h b/drivers/gpu/drm/xe/xe_device.h
> index f052c06a2d2f..4ba07617084c 100644
> --- a/drivers/gpu/drm/xe/xe_device.h
> +++ b/drivers/gpu/drm/xe/xe_device.h
> @@ -9,6 +9,7 @@
> #include <drm/drm_util.h>
>
> #include "xe_device_types.h"
> +#include "xe_sriov.h"
>
> static inline struct xe_device *to_xe_device(const struct drm_device *dev)
> {
> @@ -154,6 +155,11 @@ static inline bool xe_device_has_memirq(struct xe_device *xe)
> return GRAPHICS_VERx100(xe) >= 1250;
> }
>
> +static inline bool xe_device_needs_memirq(struct xe_device *xe)
> +{
> + return IS_SRIOV_VF(xe) && xe_device_has_memirq(xe);
> +}
maybe worth to introduce this helper earlier in the series?
> +
> u32 xe_device_ccs_bytes(struct xe_device *xe, u64 size);
>
> void xe_device_snapshot_print(struct xe_device *xe, struct drm_printer *p);
> diff --git a/drivers/gpu/drm/xe/xe_irq.c b/drivers/gpu/drm/xe/xe_irq.c
> index 18a5272d9bb0..a080d4869b16 100644
> --- a/drivers/gpu/drm/xe/xe_irq.c
> +++ b/drivers/gpu/drm/xe/xe_irq.c
> @@ -559,17 +559,15 @@ static void vf_irq_reset(struct xe_device *xe)
>
> xe_assert(xe, IS_SRIOV_VF(xe));
>
> - if (GRAPHICS_VERx100(xe) < 1210)
> - xelp_intr_disable(xe);
> - else
> + if (GRAPHICS_VERx100(xe) >= 1210) {
> xe_assert(xe, xe_device_has_memirq(xe));
> -
> - for_each_tile(tile, xe, id) {
> - if (xe_device_has_memirq(xe))
> - xe_memirq_reset(&tile->memirq);
> - else
> - gt_irq_reset(tile);
> + return;
> }
> +
> + xelp_intr_disable(xe);
> +
> + for_each_tile(tile, xe, id)
> + gt_irq_reset(tile);
> }
>
> static void xe_irq_reset(struct xe_device *xe)
> @@ -577,6 +575,11 @@ static void xe_irq_reset(struct xe_device *xe)
> struct xe_tile *tile;
> u8 id;
>
> + if (xe_device_needs_memirq(xe)) {
> + for_each_tile(tile, xe, id)
> + xe_memirq_reset(&tile->memirq);
> + }
> +
"else" or "return" ?
> if (IS_SRIOV_VF(xe))
> return vf_irq_reset(xe);
>
> @@ -604,13 +607,6 @@ static void xe_irq_reset(struct xe_device *xe)
>
> static void vf_irq_postinstall(struct xe_device *xe)
> {
> - struct xe_tile *tile;
> - unsigned int id;
> -
> - for_each_tile(tile, xe, id)
> - if (xe_device_has_memirq(xe))
> - xe_memirq_postinstall(&tile->memirq);
> -
> if (GRAPHICS_VERx100(xe) < 1210)
> xelp_intr_enable(xe, true);
> else
> @@ -619,6 +615,14 @@ static void vf_irq_postinstall(struct xe_device *xe)
>
> static void xe_irq_postinstall(struct xe_device *xe)
> {
> + struct xe_tile *tile;
> + unsigned int id;
> +
> + if (xe_device_needs_memirq(xe)) {
> + for_each_tile(tile, xe, id)
> + xe_memirq_postinstall(&tile->memirq);
> + }
> +
> if (IS_SRIOV_VF(xe))
> return vf_irq_postinstall(xe);
>
next prev parent reply other threads:[~2024-08-28 20:42 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-08-28 9:18 [PATCH v2 0/5] memirq infra changes Ilia Levi
2024-08-28 9:18 ` [PATCH v2 1/5] drm/xe: move memirq out of vf Ilia Levi
2024-08-28 20:01 ` Michal Wajdeczko
2024-08-28 9:18 ` [PATCH v2 2/5] drm/xe: memirq infra changes for msix Ilia Levi
2024-08-28 20:17 ` Michal Wajdeczko
2024-09-02 13:28 ` Ilia Levi
2024-08-28 9:18 ` [PATCH v2 3/5] drm/xe: add memirq offsets for engine instance 0 to hw engine properties Ilia Levi
2024-08-28 20:36 ` Michal Wajdeczko
2024-08-28 9:18 ` [PATCH v2 4/5] drm/xe: memirq handler changes Ilia Levi
2024-08-28 20:38 ` Michal Wajdeczko
2024-09-02 13:38 ` Ilia Levi
2024-08-28 9:18 ` [PATCH v2 5/5] drm/xe: separate memirq reset/postinstall from vf Ilia Levi
2024-08-28 20:42 ` Michal Wajdeczko [this message]
2024-09-02 13:49 ` Ilia Levi
2024-08-28 11:48 ` ✓ CI.Patch_applied: success for memirq infra changes (rev3) Patchwork
2024-08-28 11:49 ` ✓ CI.checkpatch: " Patchwork
2024-08-28 11:50 ` ✓ CI.KUnit: " Patchwork
2024-08-28 12:02 ` ✓ CI.Build: " Patchwork
2024-08-28 12:04 ` ✓ CI.Hooks: " Patchwork
2024-08-28 12:06 ` ✓ CI.checksparse: " Patchwork
2024-08-28 12:25 ` ✓ CI.BAT: " Patchwork
2024-08-28 18:53 ` ✗ CI.FULL: failure " 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=1429c1eb-91b8-43e0-9e3d-51b2367f2b9d@intel.com \
--to=michal.wajdeczko@intel.com \
--cc=ilia.levi@intel.com \
--cc=intel-xe@lists.freedesktop.org \
--cc=jonathan.cavitt@intel.com \
--cc=koby.elbaz@intel.com \
--cc=matthew.brost@intel.com \
--cc=niranjana.vishwanathapura@intel.com \
--cc=yaron.avizrat@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