From: "Lazar, Lijo" <lijo.lazar@amd.com>
To: Jiang Liu <gerry@linux.alibaba.com>,
alexander.deucher@amd.com, christian.koenig@amd.com,
Xinhui.Pan@amd.com, airlied@gmail.com, simona@ffwll.ch,
sunil.khatri@amd.com, Hawking.Zhang@amd.com,
mario.limonciello@amd.com, xiaogang.chen@amd.com,
Kent.Russell@amd.com, shuox.liu@linux.alibaba.com,
amd-gfx@lists.freedesktop.org
Subject: Re: [RFC v2 10/15] drm/admgpu: make device state machine work in stack like way
Date: Fri, 17 Jan 2025 14:24:21 +0530 [thread overview]
Message-ID: <76494817-39dd-4900-9906-e4d13639be6b@amd.com> (raw)
In-Reply-To: <f9816814dd71316bf371c01c5f4d71ca85954ade.1736732062.git.gerry@linux.alibaba.com>
On 1/13/2025 7:12 AM, Jiang Liu wrote:
> Make the device state machine work in stack like way to better support
> suspend/resume by following changes:
>
> 1. amdgpu_driver_load_kms()
> amdgpu_device_init()
> amdgpu_device_ip_early_init()
> ip_blocks[i].early_init()
> ip_blocks[i].status.valid = true
> amdgpu_device_ip_init()
> amdgpu_ras_init()
> ip_blocks[i].sw_init()
> ip_blocks[i].status.sw = true
> ip_blocks[i].hw_init()
> ip_blocks[i].status.hw = true
> amdgpu_device_ip_late_init()
> ip_blocks[i].late_init()
> ip_blocks[i].status.late_initialized = true
> amdgpu_ras_late_init()
> ras_blocks[i].ras_late_init()
> amdgpu_ras_feature_enable_on_boot()
>
> 2. amdgpu_pmops_suspend()/amdgpu_pmops_freeze()/amdgpu_pmops_poweroff()
> amdgpu_device_suspend()
> amdgpu_ras_early_fini()
> ras_blocks[i].ras_early_fini()
> amdgpu_ras_feature_disable()
> amdgpu_ras_suspend()
> amdgpu_ras_disable_all_features()
> +++ ip_blocks[i].early_fini()
> +++ ip_blocks[i].status.late_initialized = false
As said in the previous patch, please don't add confusion. You could
maintain a state machine like early fini done/late fini done etc, but
please don't introduce this kind of confusing things.
Thanks,
Lijo
> ip_blocks[i].suspend()
>
> 3. amdgpu_pmops_resume()/amdgpu_pmops_thaw()/amdgpu_pmops_restore()
> amdgpu_device_resume()
> amdgpu_device_ip_resume()
> ip_blocks[i].resume()
> amdgpu_device_ip_late_init()
> ip_blocks[i].late_init()
> ip_blocks[i].status.late_initialized = true
> amdgpu_ras_late_init()
> ras_blocks[i].ras_late_init()
> amdgpu_ras_feature_enable_on_boot()
> amdgpu_ras_resume()
> amdgpu_ras_enable_all_features()
>
> 4. amdgpu_driver_unload_kms()
> amdgpu_device_fini_hw()
> amdgpu_ras_early_fini()
> ras_blocks[i].ras_early_fini()
> +++ ip_blocks[i].early_fini()
> +++ ip_blocks[i].status.late_initialized = false
> ip_blocks[i].hw_fini()
> ip_blocks[i].status.hw = false
>
> 5. amdgpu_driver_release_kms()
> amdgpu_device_fini_sw()
> amdgpu_device_ip_fini()
> ip_blocks[i].sw_fini()
> ip_blocks[i].status.sw = false
> --- ip_blocks[i].status.valid = false
> +++ amdgpu_ras_fini()
> ip_blocks[i].late_fini()
> +++ ip_blocks[i].status.valid = false
> --- ip_blocks[i].status.late_initialized = false
> --- amdgpu_ras_fini()
>
> The main changes include:
> 1) invoke ip_blocks[i].early_fini in amdgpu_pmops_suspend().
> 2) set ip_blocks[i].status.late_initialized to false after calling
> callback `early_fini`. We have auditted all usages of the
> late_initialized flag and no functional changes found.
> 3) only set ip_blocks[i].status.valid = false after calling the
> `late_fini` callback.
> 4) call amdgpu_ras_fini() before invoking ip_blocks[i].late_fini.
>
> There's one more task left to analyze GPU reset related state machine
> transitions.
>
> Signed-off-by: Jiang Liu <gerry@linux.alibaba.com>
> ---
> drivers/gpu/drm/amd/amdgpu/amdgpu_device.c | 22 ++++++++++++++++++++--
> 1 file changed, 20 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> index 6b503fb7e366..c2e4057ecd82 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_device.c
> @@ -3449,6 +3449,8 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
> adev->ip_blocks[i].status.sw = false;
> }
>
> + amdgpu_ras_fini(adev);
> +
> for (i = adev->num_ip_blocks - 1; i >= 0; i--) {
> if (!adev->ip_blocks[i].status.valid)
> continue;
> @@ -3457,8 +3459,6 @@ static int amdgpu_device_ip_fini(struct amdgpu_device *adev)
> adev->ip_blocks[i].status.valid = false;
> }
>
> - amdgpu_ras_fini(adev);
> -
> return 0;
> }
>
> @@ -3516,6 +3516,24 @@ static int amdgpu_device_ip_suspend_phase1(struct amdgpu_device *adev)
> if (amdgpu_dpm_set_df_cstate(adev, DF_CSTATE_DISALLOW))
> dev_warn(adev->dev, "Failed to disallow df cstate");
>
> + for (i = adev->num_ip_blocks - 1; i >= 0; i--) {
> + if (!adev->ip_blocks[i].status.valid)
> + continue;
> + if (!adev->ip_blocks[i].status.late_initialized)
> + continue;
> +
> + if (adev->ip_blocks[i].version->funcs->early_fini) {
> + r = adev->ip_blocks[i].version->funcs->early_fini(&adev->ip_blocks[i]);
> + if (r) {
> + DRM_ERROR(" of IP block <%s> failed %d\n",
> + adev->ip_blocks[i].version->funcs->name, r);
> + return r;
> + }
> + }
> +
> + adev->ip_blocks[i].status.late_initialized = false;
> + }
> +
> for (i = adev->num_ip_blocks - 1; i >= 0; i--) {
> if (!adev->ip_blocks[i].status.valid)
> continue;
next prev parent reply other threads:[~2025-01-17 8:54 UTC|newest]
Thread overview: 39+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-01-13 1:42 [RFC v2 00/15] Enhance device state machine to better support suspend/resume Jiang Liu
2025-01-13 1:42 ` [RFC v2 01/15] drm/amdgpu: add helper functions to track status for ras manager Jiang Liu
2025-01-17 1:13 ` Wang, Yang(Kevin)
2025-01-17 5:03 ` Lazar, Lijo
2025-01-13 1:42 ` [RFC v2 02/15] drm/amdgpu: add a flag to track ras debugfs creation status Jiang Liu
2025-01-17 5:24 ` Lazar, Lijo
2025-01-13 1:42 ` [RFC v2 03/15] drm/amdgpu: free all resources on error recovery path of amdgpu_ras_init() Jiang Liu
2025-01-16 21:02 ` Mario Limonciello
2025-01-17 5:39 ` Lazar, Lijo
2025-01-13 1:42 ` [RFC v2 04/15] drm/amdgpu: introduce a flag to track refcount held for features Jiang Liu
2025-01-17 5:46 ` Lazar, Lijo
2025-01-13 1:42 ` [RFC v2 05/15] drm/amdgpu: enhance amdgpu_ras_block_late_fini() Jiang Liu
2025-01-16 21:10 ` Mario Limonciello
2025-01-17 5:54 ` Lazar, Lijo
2025-01-13 1:42 ` [RFC v2 06/15] drm/amdgpu: enhance amdgpu_ras_pre_fini() to better support SR Jiang Liu
2025-01-16 21:19 ` Mario Limonciello
2025-01-17 6:09 ` Lazar, Lijo
2025-01-13 1:42 ` [RFC v2 07/15] drm/admgpu: rename amdgpu_ras_pre_fini() to amdgpu_ras_early_fini() Jiang Liu
2025-01-16 21:25 ` Mario Limonciello
2025-01-17 1:19 ` Wang, Yang(Kevin)
2025-01-17 8:37 ` Lazar, Lijo
2025-01-13 1:42 ` [RFC v2 08/15] drm/amdgpu: make IP block state machine works in stack like way Jiang Liu
2025-01-17 8:45 ` Lazar, Lijo
2025-01-13 1:42 ` [RFC v2 09/15] drm/amdgpu_dm: enhance amdgpu_dm_early_fini() for PM ops Jiang Liu
2025-01-16 21:30 ` Mario Limonciello
2025-01-13 1:42 ` [RFC v2 10/15] drm/admgpu: make device state machine work in stack like way Jiang Liu
2025-01-13 22:27 ` Mario Limonciello
2025-01-14 1:58 ` Gerry Liu
2025-01-15 19:36 ` Mario Limonciello
2025-01-17 8:54 ` Lazar, Lijo [this message]
2025-01-13 1:42 ` [RFC v2 11/15] drm/amdgpu: convert ip block bool flags into an enum Jiang Liu
2025-01-17 8:57 ` Lazar, Lijo
2025-01-13 1:42 ` [RFC v2 12/15] drm/amdgpu: introduce IP block iterators to reduce duplicated code Jiang Liu
2025-01-13 1:42 ` [RFC v2 13/15] drm/amdgpu: walk IP blocks in reverse order when shutdown Jiang Liu
2025-01-13 22:28 ` Mario Limonciello
2025-01-13 1:42 ` [RFC v2 14/15] drm/amdgpu/nbio: improve the way to manage irq reference count Jiang Liu
2025-01-13 1:42 ` [RFC v2 15/15] drm/amdgpu/asic: make ip block operations symmetric by .early_fini() Jiang Liu
2025-01-20 6:27 ` [RFC v2 00/15] Enhance device state machine to better support suspend/resume Zhang, Hawking
2025-01-23 0:02 ` Mika Laitio
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=76494817-39dd-4900-9906-e4d13639be6b@amd.com \
--to=lijo.lazar@amd.com \
--cc=Hawking.Zhang@amd.com \
--cc=Kent.Russell@amd.com \
--cc=Xinhui.Pan@amd.com \
--cc=airlied@gmail.com \
--cc=alexander.deucher@amd.com \
--cc=amd-gfx@lists.freedesktop.org \
--cc=christian.koenig@amd.com \
--cc=gerry@linux.alibaba.com \
--cc=mario.limonciello@amd.com \
--cc=shuox.liu@linux.alibaba.com \
--cc=simona@ffwll.ch \
--cc=sunil.khatri@amd.com \
--cc=xiaogang.chen@amd.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