[PATCH] drm/amdkfd: correct sdma queue number in kfd device init
Kim, Jonathan
Jonathan.Kim at amd.com
Fri Dec 17 15:34:50 UTC 2021
> -----Original Message-----
> From: Sider, Graham <Graham.Sider at amd.com>
> Sent: December 17, 2021 10:06 AM
> To: Chen, Guchun <Guchun.Chen at amd.com>; amd-
> gfx at lists.freedesktop.org; Deucher, Alexander
> <Alexander.Deucher at amd.com>; Kuehling, Felix
> <Felix.Kuehling at amd.com>; Kim, Jonathan <Jonathan.Kim at amd.com>
> Subject: RE: [PATCH] drm/amdkfd: correct sdma queue number in kfd
> device init
>
> [Public]
>
> > -----Original Message-----
> > From: Chen, Guchun <Guchun.Chen at amd.com>
> > Sent: Friday, December 17, 2021 9:31 AM
> > To: amd-gfx at lists.freedesktop.org; Deucher, Alexander
> > <Alexander.Deucher at amd.com>; Sider, Graham
> <Graham.Sider at amd.com>;
> > Kuehling, Felix <Felix.Kuehling at amd.com>; Kim, Jonathan
> > <Jonathan.Kim at amd.com>
> > Cc: Chen, Guchun <Guchun.Chen at amd.com>
> > Subject: [PATCH] drm/amdkfd: correct sdma queue number in kfd device
> > init
> >
> > sdma queue number is not correct like on vega20, this patch promises
> > the setting keeps the same after code refactor.
> > Additionally, improve code to use switch case to list IP version to
> > complete kfd device_info structure filling.
> > This keeps consistency with the IP parse code in amdgpu_discovery.c.
> >
> > Fixes: a9e2c4dc6cc4("drm/amdkfd: add kfd_device_info_init function")
> > Signed-off-by: Guchun Chen <guchun.chen at amd.com>
> > ---
> > drivers/gpu/drm/amd/amdkfd/kfd_device.c | 74
> > ++++++++++++++++++++++---
> > 1 file changed, 65 insertions(+), 9 deletions(-)
> >
> > diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
> > b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
> > index facc28f58c1f..e50bf992f298 100644
> > --- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
> > +++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
> > @@ -59,11 +59,72 @@ static void kfd_gtt_sa_fini(struct kfd_dev *kfd);
> >
> > static int kfd_resume(struct kfd_dev *kfd);
> >
> > +static void kfd_device_info_set_sdma_queue_num(struct kfd_dev *kfd)
> {
> > + uint32_t sdma_version = kfd->adev-
> >ip_versions[SDMA0_HWIP][0];
> > +
> > + switch (sdma_version) {
> > + case IP_VERSION(4, 0, 0):/* VEGA10 */
> > + case IP_VERSION(4, 0, 1):/* VEGA12 */
> > + case IP_VERSION(4, 1, 0):/* RAVEN */
> > + case IP_VERSION(4, 1, 1):/* RAVEN */
> > + case IP_VERSION(4, 1, 2):/* RENIOR */
> > + case IP_VERSION(5, 2, 1):/* VANGOGH */
> > + case IP_VERSION(5, 2, 3):/* YELLOW_CARP */
> > + kfd->device_info.num_sdma_queues_per_engine =
> > 2;
> > + break;
> > + case IP_VERSION(4, 2, 0):/* VEGA20 */
>
> Thanks for spotting this Guchun. My previous patch should have used a "<"
> instead of a "<=" on IP_VERSION(4, 2, 0).
>
> > + case IP_VERSION(4, 2, 2):/* ARCTUTUS */
> > + case IP_VERSION(4, 4, 0):/* ALDEBARAN */
> > + case IP_VERSION(5, 0, 0):/* NAVI10 */
> > + case IP_VERSION(5, 0, 1):/* CYAN_SKILLFISH */
> > + case IP_VERSION(5, 0, 2):/* NAVI14 */
> > + case IP_VERSION(5, 0, 5):/* NAVI12 */
> > + case IP_VERSION(5, 2, 0):/* SIENNA_CICHLID */
> > + case IP_VERSION(5, 2, 2):/* NAVY_FLOUDER */
> > + case IP_VERSION(5, 2, 4):/* DIMGREY_CAVEFISH */
> > + kfd->device_info.num_sdma_queues_per_engine =
> > 8;
> > + break;
> > + default:
> > + dev_err(kfd_device,
> > + "Failed to find sdma ip
> > blocks(SDMA_HWIP:0x%x) in %s\n",
> > + sdma_version, __func__);
> > + }
> > +}
> > +
> > +static void kfd_device_info_set_event_interrupt_class(struct kfd_dev
> > +*kfd) {
> > + uint32_t gc_version = KFD_GC_VERSION(kfd);
> > +
> > + switch (gc_version) {
> > + case IP_VERSION(9, 0, 1): /* VEGA10 */
> > + case IP_VERSION(9, 2, 1): /* VEGA12 */
> > + case IP_VERSION(9, 3, 0): /* RENOIR */
> > + case IP_VERSION(9, 4, 0): /* VEGA20 */
> > + case IP_VERSION(9, 4, 1): /* ARCTURUS */
> > + case IP_VERSION(9, 4, 2): /* ALDEBARAN */
> > + case IP_VERSION(10, 3, 1): /* VANGOGH */
> > + case IP_VERSION(10, 3, 3): /* YELLOW_CARP */
> > + case IP_VERSION(10, 1, 3): /* CYAN_SKILLFISH */
> > + case IP_VERSION(10, 1, 10): /* NAVI10 */
> > + case IP_VERSION(10, 1, 2): /* NAVI12 */
> > + case IP_VERSION(10, 1, 1): /* NAVI14 */
> > + case IP_VERSION(10, 3, 0): /* SIENNA_CICHLID */
> > + case IP_VERSION(10, 3, 2): /* NAVY_FLOUNDER */
> > + case IP_VERSION(10, 3, 4): /* DIMGREY_CAVEFISH */
> > + case IP_VERSION(10, 3, 5): /* BEIGE_GOBY */
> > + kfd->device_info.event_interrupt_class =
> > &event_interrupt_class_v9;
> > + break;
> > + default:
> > + dev_err(kfd_device, "Failed to find gc ip
> > blocks(GC_HWIP:0x%x) in %s\n",
> > + gc_version, __func__);
> > + }
> > +}
>
> I understand the appeal of moving to a switch for the SDMA queue num
> above since its setting isn't very linear w.r.t. the SDMA versioning. That said
> I don't know if I understand moving to a switch for the event interrupt class
> here. To clarify, originally when I set all SOC15 to event_interrupt_class_v9
> it was to follow what was in the device_info structs in drm-staging-next at
> that time--that said if the plan is in a following patch to change this such
> that gfx10 are set to to event_interrupt_class_v10, what's the reasoning
> not to format it along the lines of:
>
> if (gc_version >= IP_VERSION(10, 1, 1)
> kfd->device_info.event_interrupt_class =
> &event_interrupt_class_v10; else
> kfd->device_info.event_interrupt_class =
> &event_interrupt_class_v9;
>
> (Assuming this is done in the SOC15 case, and of course cases would need
> to be added here eventually for e.g. event_interrupt_class_v11, but not
> necessarily for every asic).
Explicit hard checks with a non-assignment on default might have advantages by not allowing the KFD to proceed to load for unregistered devices or force developers to assign the correct interrupt class without making assumptions.
But that means more maintenance and additional handling on non-assignment cases to fail gracefully.
Thanks,
Jon
>
> Best,
> Graham
>
> > +
> > static void kfd_device_info_init(struct kfd_dev *kfd,
> > bool vf, uint32_t gfx_target_version) {
> > uint32_t gc_version = KFD_GC_VERSION(kfd);
> > - uint32_t sdma_version = kfd->adev-
> >ip_versions[SDMA0_HWIP][0];
> > uint32_t asic_type = kfd->adev->asic_type;
> >
> > kfd->device_info.max_pasid_bits = 16; @@ -75,16 +136,11 @@
> static
> > void kfd_device_info_init(struct kfd_dev *kfd,
> > if (KFD_IS_SOC15(kfd)) {
> > kfd->device_info.doorbell_size = 8;
> > kfd->device_info.ih_ring_entry_size = 8 * sizeof(uint32_t);
> > - kfd->device_info.event_interrupt_class =
> > &event_interrupt_class_v9;
> > kfd->device_info.supports_cwsr = true;
> >
> > - if ((sdma_version >= IP_VERSION(4, 0, 0) &&
> > - sdma_version <= IP_VERSION(4, 2, 0)) ||
> > - sdma_version == IP_VERSION(5, 2, 1) ||
> > - sdma_version == IP_VERSION(5, 2, 3))
> > - kfd->device_info.num_sdma_queues_per_engine =
> > 2;
> > - else
> > - kfd->device_info.num_sdma_queues_per_engine =
> > 8;
> > + kfd_device_info_set_sdma_queue_num(kfd);
> > +
> > + kfd_device_info_set_event_interrupt_class(kfd);
> >
> > /* Raven */
> > if (gc_version == IP_VERSION(9, 1, 0) ||
> > --
> > 2.17.1
More information about the amd-gfx
mailing list