[Intel-gfx] [PATCH v4 1/3] drm/i915: Introduce new macros for i915 PTE
Michael Cheng
michael.cheng at intel.com
Sat Nov 13 01:42:28 UTC 2021
Thanks for the feed back! I feel like using something name GEN6 or BYT
for a platform that's not GEN6 or BYT could be a bit confusing, that's
why we decided to go with something more generic. I do agree I need to
cite the bspec more. Ill wait for more feedback before I send a new
revision out.
On 2021-11-12 5:31 p.m., Matt Roper wrote:
> On Fri, Nov 12, 2021 at 05:28:09PM -0800, Matt Roper wrote:
>> On Wed, Nov 10, 2021 at 04:45:47PM -0800, Michael Cheng wrote:
>>> Certain functions within i915 uses macros that are defined for
>>> specific architectures by the mmu, such as _PAGE_RW and _PAGE_PRESENT
>>> (Some architectures don't even have these macros defined, like ARM64).
>>>
>>> Instead of re-using bits defined for the CPU, we should use bits
>>> defined for i915. This patch introduces two new macros,
>>> I915_PAGE_PRESENT and I915_PAGE_RW, to check for bits 0 and 1 and, to
>>> replace all occurrences of _PAGE_RW and _PAGE_PRESENT within i915.
>> On older platforms we already had our own definition of GEN6_PTE_VALID
>> (the spec uses "present" and "valid" interchangeably) which we were
>> using to encode our ggtt ptes up through HSW; it might be better to go
>> back to using that rather than adding a new define.
>>
>> It looks like BYT is when the writable bit showed up, and we did add a
>> new define there (BYT_PTE_WRITEABLE), but on the next platform (BDW) we
>> switched over to using the CPU page table flags instead and never used
>> it again. So we could probably replace _PAGE_RW with BYT_PTE_WRITEABLE
>> as well.
> Okay, I should have looked at the rest of the series before reviewing
> the first patch; it looks like your next two patches replace
> GEN6_PTE_VALID and BYT_PTE_WRITEABLE with the new definitions here. I
> still think it might be preferable to reuse the existing macros (which
> also help clarify the platforms that those bits first showed up in the
> PTE on) rather than replacing them with new macros, but I don't feel
> super strongly about it if other reviewers feel differently.
>
>
> Matt
>
>>> Looking at the bspecs for pre gen 12 and gen 12, bit 0 and 1 are the
>>> same throughout the generations.
>> This last sentence seems a bit confusing --- it's true that bit 0 has
>> always been a present/valid flag, but bit 1 wasn't a writable bit until
>> BYT; there just wasn't a writable bit at all (e.g., bspec page 229).
>>
>> It might be worth tossing a few bspec references on the commit message
>> here, just for future reference. E.g.,
>>
>> Bspec: 253, 45039
>>
>>
>> Matt
>>
>>> Signed-off-by: Michael Cheng <michael.cheng at intel.com>
>>> ---
>>> drivers/gpu/drm/i915/gt/gen8_ppgtt.c | 6 +++---
>>> drivers/gpu/drm/i915/gt/intel_ggtt.c | 2 +-
>>> drivers/gpu/drm/i915/gt/intel_gtt.h | 3 +++
>>> drivers/gpu/drm/i915/gvt/gtt.c | 12 ++++++------
>>> 4 files changed, 13 insertions(+), 10 deletions(-)
>>>
>>> diff --git a/drivers/gpu/drm/i915/gt/gen8_ppgtt.c b/drivers/gpu/drm/i915/gt/gen8_ppgtt.c
>>> index 9966e9dc5218..f89b50ffc286 100644
>>> --- a/drivers/gpu/drm/i915/gt/gen8_ppgtt.c
>>> +++ b/drivers/gpu/drm/i915/gt/gen8_ppgtt.c
>>> @@ -18,7 +18,7 @@
>>> static u64 gen8_pde_encode(const dma_addr_t addr,
>>> const enum i915_cache_level level)
>>> {
>>> - u64 pde = addr | _PAGE_PRESENT | _PAGE_RW;
>>> + u64 pde = addr | I915_PAGE_PRESENT | I915_PAGE_RW;
>>>
>>> if (level != I915_CACHE_NONE)
>>> pde |= PPAT_CACHED_PDE;
>>> @@ -32,10 +32,10 @@ static u64 gen8_pte_encode(dma_addr_t addr,
>>> enum i915_cache_level level,
>>> u32 flags)
>>> {
>>> - gen8_pte_t pte = addr | _PAGE_PRESENT | _PAGE_RW;
>>> + gen8_pte_t pte = addr | I915_PAGE_PRESENT | I915_PAGE_RW;
>>>
>>> if (unlikely(flags & PTE_READ_ONLY))
>>> - pte &= ~_PAGE_RW;
>>> + pte &= ~I915_PAGE_RW;
>>>
>>> if (flags & PTE_LM)
>>> pte |= GEN12_PPGTT_PTE_LM;
>>> diff --git a/drivers/gpu/drm/i915/gt/intel_ggtt.c b/drivers/gpu/drm/i915/gt/intel_ggtt.c
>>> index 1fb4a03d7ac3..3f8e1ee0fbfa 100644
>>> --- a/drivers/gpu/drm/i915/gt/intel_ggtt.c
>>> +++ b/drivers/gpu/drm/i915/gt/intel_ggtt.c
>>> @@ -207,7 +207,7 @@ u64 gen8_ggtt_pte_encode(dma_addr_t addr,
>>> enum i915_cache_level level,
>>> u32 flags)
>>> {
>>> - gen8_pte_t pte = addr | _PAGE_PRESENT;
>>> + gen8_pte_t pte = addr | I915_PAGE_PRESENT;
>>>
>>> if (flags & PTE_LM)
>>> pte |= GEN12_GGTT_PTE_LM;
>>> diff --git a/drivers/gpu/drm/i915/gt/intel_gtt.h b/drivers/gpu/drm/i915/gt/intel_gtt.h
>>> index dfeaef680aac..fba9c0c18f4a 100644
>>> --- a/drivers/gpu/drm/i915/gt/intel_gtt.h
>>> +++ b/drivers/gpu/drm/i915/gt/intel_gtt.h
>>> @@ -39,6 +39,9 @@
>>>
>>> #define NALLOC 3 /* 1 normal, 1 for concurrent threads, 1 for preallocation */
>>>
>>> +#define I915_PAGE_PRESENT BIT_ULL(0)
>>> +#define I915_PAGE_RW BIT_ULL(1)
>>> +
>>> #define I915_GTT_PAGE_SIZE_4K BIT_ULL(12)
>>> #define I915_GTT_PAGE_SIZE_64K BIT_ULL(16)
>>> #define I915_GTT_PAGE_SIZE_2M BIT_ULL(21)
>>> diff --git a/drivers/gpu/drm/i915/gvt/gtt.c b/drivers/gpu/drm/i915/gvt/gtt.c
>>> index 53d0cb327539..8f6a055854f7 100644
>>> --- a/drivers/gpu/drm/i915/gvt/gtt.c
>>> +++ b/drivers/gpu/drm/i915/gvt/gtt.c
>>> @@ -446,17 +446,17 @@ static bool gen8_gtt_test_present(struct intel_gvt_gtt_entry *e)
>>> || e->type == GTT_TYPE_PPGTT_ROOT_L4_ENTRY)
>>> return (e->val64 != 0);
>>> else
>>> - return (e->val64 & _PAGE_PRESENT);
>>> + return (e->val64 & I915_PAGE_PRESENT);
>>> }
>>>
>>> static void gtt_entry_clear_present(struct intel_gvt_gtt_entry *e)
>>> {
>>> - e->val64 &= ~_PAGE_PRESENT;
>>> + e->val64 &= ~I915_PAGE_PRESENT;
>>> }
>>>
>>> static void gtt_entry_set_present(struct intel_gvt_gtt_entry *e)
>>> {
>>> - e->val64 |= _PAGE_PRESENT;
>>> + e->val64 |= I915_PAGE_PRESENT;
>>> }
>>>
>>> static bool gen8_gtt_test_64k_splited(struct intel_gvt_gtt_entry *e)
>>> @@ -2439,7 +2439,7 @@ static int alloc_scratch_pages(struct intel_vgpu *vgpu,
>>> /* The entry parameters like present/writeable/cache type
>>> * set to the same as i915's scratch page tree.
>>> */
>>> - se.val64 |= _PAGE_PRESENT | _PAGE_RW;
>>> + se.val64 |= I915_PAGE_PRESENT | I915_PAGE_RW;
>>> if (type == GTT_TYPE_PPGTT_PDE_PT)
>>> se.val64 |= PPAT_CACHED;
>>>
>>> @@ -2896,7 +2896,7 @@ void intel_gvt_restore_ggtt(struct intel_gvt *gvt)
>>> offset = vgpu_aperture_gmadr_base(vgpu) >> PAGE_SHIFT;
>>> for (idx = 0; idx < num_low; idx++) {
>>> pte = mm->ggtt_mm.host_ggtt_aperture[idx];
>>> - if (pte & _PAGE_PRESENT)
>>> + if (pte & I915_PAGE_PRESENT)
>>> write_pte64(vgpu->gvt->gt->ggtt, offset + idx, pte);
>>> }
>>>
>>> @@ -2904,7 +2904,7 @@ void intel_gvt_restore_ggtt(struct intel_gvt *gvt)
>>> offset = vgpu_hidden_gmadr_base(vgpu) >> PAGE_SHIFT;
>>> for (idx = 0; idx < num_hi; idx++) {
>>> pte = mm->ggtt_mm.host_ggtt_hidden[idx];
>>> - if (pte & _PAGE_PRESENT)
>>> + if (pte & I915_PAGE_PRESENT)
>>> write_pte64(vgpu->gvt->gt->ggtt, offset + idx, pte);
>>> }
>>> }
>>> --
>>> 2.25.1
>>>
>> --
>> Matt Roper
>> Graphics Software Engineer
>> VTT-OSGC Platform Enablement
>> Intel Corporation
>> (916) 356-2795
More information about the Intel-gfx
mailing list