From: Nirmoy Das <nirmoy.das@intel.com>
To: "Souza, Jose" <jose.souza@intel.com>,
"intel-xe@lists.freedesktop.org" <intel-xe@lists.freedesktop.org>
Subject: Re: [PATCH v2 5/7] drm/xe: Introduce has_device_atomics_on_smem device info
Date: Fri, 12 Apr 2024 17:22:33 +0200 [thread overview]
Message-ID: <226bbca6-35ac-452f-a86d-badff5bd56ca@intel.com> (raw)
In-Reply-To: <2fc9b73615bb9fb5115773fcc6e4fc9243a6d234.camel@intel.com>
Hi Jose,
On 4/12/2024 3:31 PM, Souza, Jose wrote:
> On Fri, 2024-04-12 at 11:22 +0200, Nirmoy Das wrote:
>> Add has_device_atomics_on_smem to specify that a device
>> supports device atomics on system memory. Currently XE2
>> supports this so set this for XE2.
>>
>> Signed-off-by: Nirmoy Das <nirmoy.das@intel.com>
>> ---
>> drivers/gpu/drm/xe/xe_device_types.h | 2 ++
>> drivers/gpu/drm/xe/xe_pci.c | 2 ++
>> drivers/gpu/drm/xe/xe_pci_types.h | 1 +
>> 3 files changed, 5 insertions(+)
>>
>> diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h
>> index bde42fe288ff..c06be1e00f1a 100644
>> --- a/drivers/gpu/drm/xe/xe_device_types.h
>> +++ b/drivers/gpu/drm/xe/xe_device_types.h
>> @@ -283,6 +283,8 @@ struct xe_device {
>> u8 skip_guc_pc:1;
>> /** @info.has_atomic_enable_pte_bit: Device has atomic enable PTE bit */
>> u8 has_atomic_enable_pte_bit:1;
>> + /** @info.has_device_atomics_on_smem: Supports device atomics on SMEM */
>> + u8 has_device_atomics_on_smem:1;
>>
> This one can be easily replaced by a simple gfx_ip >= 20.
>
> In my opinion it should only be added to xe_graphics_desc when feature check is more complex,
Did occur to me but I haven't seen any existing example so picked a safe
option. I will add platform check in
xe_info_init() if no one objects.
Thanks,
Nirmoy
> like the has_atomic_enable_pte_bit.
> Otherwise it will get complex and error prone to add new platforms.
>
>> #if IS_ENABLED(CONFIG_DRM_XE_DISPLAY)
>> struct {
>> diff --git a/drivers/gpu/drm/xe/xe_pci.c b/drivers/gpu/drm/xe/xe_pci.c
>> index 9cf45559b915..4e4699962337 100644
>> --- a/drivers/gpu/drm/xe/xe_pci.c
>> +++ b/drivers/gpu/drm/xe/xe_pci.c
>> @@ -165,6 +165,7 @@ static const struct xe_graphics_desc graphics_xelpg = {
>> .dma_mask_size = 46, \
>> .has_asid = 1, \
>> .has_atomic_enable_pte_bit = 1, \
>> + .has_device_atomics_on_smem = 1, \
>> .has_flat_ccs = 1, \
>> .has_range_tlb_invalidation = 1, \
>> .has_usm = 1, \
>> @@ -630,6 +631,7 @@ static int xe_info_init(struct xe_device *xe,
>> xe->info.vm_max_level = graphics_desc->vm_max_level;
>> xe->info.has_asid = graphics_desc->has_asid;
>> xe->info.has_atomic_enable_pte_bit = graphics_desc->has_atomic_enable_pte_bit;
>> + xe->info.has_device_atomics_on_smem = graphics_desc->has_device_atomics_on_smem;
>> xe->info.has_flat_ccs = graphics_desc->has_flat_ccs;
>> xe->info.has_range_tlb_invalidation = graphics_desc->has_range_tlb_invalidation;
>> xe->info.has_usm = graphics_desc->has_usm;
>> diff --git a/drivers/gpu/drm/xe/xe_pci_types.h b/drivers/gpu/drm/xe/xe_pci_types.h
>> index e1f2b4879fc2..d403814a7a84 100644
>> --- a/drivers/gpu/drm/xe/xe_pci_types.h
>> +++ b/drivers/gpu/drm/xe/xe_pci_types.h
>> @@ -26,6 +26,7 @@ struct xe_graphics_desc {
>>
>> u8 has_asid:1;
>> u8 has_atomic_enable_pte_bit:1;
>> + u8 has_device_atomics_on_smem:1;
>> u8 has_flat_ccs:1;
>> u8 has_range_tlb_invalidation:1;
>> u8 has_usm:1;
next prev parent reply other threads:[~2024-04-12 15:22 UTC|newest]
Thread overview: 21+ messages / expand[flat|nested] mbox.gz Atom feed top
2024-04-12 9:22 [PATCH v2 0/7] Enable device atomics with a VM bind flag Nirmoy Das
2024-04-12 9:22 ` [PATCH v2 1/7] drm/xe: Introduce has_atomic_enable_pte_bit device info Nirmoy Das
2024-04-12 9:22 ` [PATCH v2 2/7] drm/xe: Consolidate setting PTE_AE into one place Nirmoy Das
2024-04-12 9:22 ` [PATCH v2 3/7] drm/xe: Add function to check if BO has single placement Nirmoy Das
2024-04-12 9:22 ` [PATCH v2 4/7] drm/xe: Move vm bind bo validation to a helper function Nirmoy Das
2024-04-12 18:37 ` Matthew Brost
2024-04-12 9:22 ` [PATCH v2 5/7] drm/xe: Introduce has_device_atomics_on_smem device info Nirmoy Das
2024-04-12 13:31 ` Souza, Jose
2024-04-12 15:22 ` Nirmoy Das [this message]
2024-04-12 9:22 ` [PATCH v2 6/7] drm/xe/uapi: Introduce VMA bind flag for device atomics Nirmoy Das
2024-04-12 9:22 ` [PATCH v2 7/7] drm/xe/uapi: Add a query flag for has_device_atomics_on_smem Nirmoy Das
2024-04-12 13:37 ` Souza, Jose
2024-04-12 15:16 ` Nirmoy Das
2024-04-12 9:58 ` ✓ CI.Patch_applied: success for Enable device atomics with a VM bind flag (rev2) Patchwork
2024-04-12 9:58 ` ✓ CI.checkpatch: " Patchwork
2024-04-12 9:59 ` ✓ CI.KUnit: " Patchwork
2024-04-12 10:11 ` ✓ CI.Build: " Patchwork
2024-04-12 10:13 ` ✓ CI.Hooks: " Patchwork
2024-04-12 10:15 ` ✓ CI.checksparse: " Patchwork
2024-04-12 10:38 ` ✗ CI.BAT: failure " Patchwork
2024-04-15 7:52 ` ✗ CI.FULL: " 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=226bbca6-35ac-452f-a86d-badff5bd56ca@intel.com \
--to=nirmoy.das@intel.com \
--cc=intel-xe@lists.freedesktop.org \
--cc=jose.souza@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