Skip to contentSkip to conversation

powerpc/powernv/npu: Remove redundant change_pte() hook

PATCH Applied Peter Xu 5 participants 5 messages Jan 31, 2019 – Feb 22, 2019 PowerPC

lore
Share & tools
lore

Peter Xu

author
#0 lore raw

PATCH 1/1 powerpc/powernv/npu: Remove redundant change_pte() hook

The change_pte() notifier was designed to use as a quick path to update secondary MMU PTEs on write permission changes or PFN changes. For KVM, it could reduce the vm-exits when vcpu faults on the pages that was touched up by KSM. It's not used to do cache invalidations, for example, if we see the notifier will be called before the real PTE update after all (please see set_pte_at_notify that set_pte_at was called later).

All the necessary cache invalidation should all be done in invalidate_range() already.

CC Benjamin Herrenschmidt <[email protected]>
CC Paul Mackerras <[email protected]>
CC Michael Ellerman <[email protected]>
CC Alistair Popple <[email protected]>
CC Alexey Kardashevskiy <[email protected]>
CC Mark Hairgrove <[email protected]>
CC Balbir Singh <[email protected]>
CC David Gibson <[email protected]>
CC Andrea Arcangeli <[email protected]>
CC Jerome Glisse <[email protected]>
CC Jason Wang <[email protected]>
---
 arch/powerpc/platforms/powernv/npu-dma.c | 10 ----------
 1 file changed, 10 deletions(-)
arch/powerpc/platforms/powernv/npu-dma.c +0 -10
index 3f58c7dbd581..c003b29d870e 100644
@@ -917,15 +917,6 @@ static void pnv_npu2_mn_release(struct mmu_notifier *mn,
 	mmio_invalidate(npu_context, 0, ~0UL);
 }
 
-static void pnv_npu2_mn_change_pte(struct mmu_notifier *mn,
-				struct mm_struct *mm,
-				unsigned long address,
-				pte_t pte)
-{
-	struct npu_context *npu_context = mn_to_npu_context(mn);
-	mmio_invalidate(npu_context, address, PAGE_SIZE);
-}
-
 static void pnv_npu2_mn_invalidate_range(struct mmu_notifier *mn,
 					struct mm_struct *mm,
 					unsigned long start, unsigned long end)
@@ -936,7 +927,6 @@ static void pnv_npu2_mn_invalidate_range(struct mmu_notifier *mn,
 
 static const struct mmu_notifier_ops nv_nmmu_notifier_ops = {
 	.release = pnv_npu2_mn_release,
-	.change_pte = pnv_npu2_mn_change_pte,
 	.invalidate_range = pnv_npu2_mn_invalidate_range,
 };
 
Signature · 1 line
2.17.1
Reviewed-by on v1 1/1
#1 lore raw
Quoted 25 lines · Peter Xu

The change_pte() notifier was designed to use as a quick path to update secondary MMU PTEs on write permission changes or PFN changes. For KVM, it could reduce the vm-exits when vcpu faults on the pages that was touched up by KSM. It's not used to do cache invalidations, for example, if we see the notifier will be called before the real PTE update after all (please see set_pte_at_notify that set_pte_at was called later).

All the necessary cache invalidation should all be done in invalidate_range() already.

CC Benjamin Herrenschmidt <[email protected]>
CC Paul Mackerras <[email protected]>
CC Michael Ellerman <[email protected]>
CC Alistair Popple <[email protected]>
CC Alexey Kardashevskiy <[email protected]>
CC Mark Hairgrove <[email protected]>
CC Balbir Singh <[email protected]>
CC David Gibson <[email protected]>
CC Andrea Arcangeli <[email protected]>
CC Jerome Glisse <[email protected]>
CC Jason Wang <[email protected]>
Quoted 3 lines
---
 arch/powerpc/platforms/powernv/npu-dma.c | 10 ----------
 1 file changed, 10 deletions(-)

It doesn't make sense to implement change_pte as an invalidate, change_pte is not compulsory to implement so if one wants to have invalidates only, change_pte method shouldn't be implemented in the first place and the common code will guarantee to invoke the range invalidates instead.

Currently the whole change_pte optimization is effectively disabled as noted in past discussions with Jerome (because of the range invalidates that always surrounds it), so we need to revisit the whole change_pte logic and decide it to re-enable it or to drop it as a whole, but in the meantime it's good to cleanup spots like below that should leave change_pte alone.

There are several examples of mmu_notifiers_ops in the kernel that don't implement change_pte, in fact it's the majority. Of all mmu notifier users, only nv_nmmu_notifier_ops, intel_mmuops_change and kvm_mmu_notifier_ops implements change_pte and as Peter found out by source review nv_nmmu_notifier_ops, intel_mmuops_change are wrong about it and should stop implementing it as an invalidate.

In short change_pte is only implemented correctly from KVM which can really updates the spte and flushes the TLB but the spte update remains and could avoid a vmexit if we figure out how to re-enable the optimization safely (the TLB fill after change_pte in KVM EPT/shadow secondary MMU will be looked up by the CPU in hardware).

If change_pte is implemented, it should update the mapping like KVM does and not do an invalidate.

Thanks,
Andrea

Quoted 32 lines
arch/powerpc/platforms/powernv/npu-dma.c +0 -10
index 3f58c7dbd581..c003b29d870e 100644
@@ -917,15 +917,6 @@ static void pnv_npu2_mn_release(struct mmu_notifier *mn,
 	mmio_invalidate(npu_context, 0, ~0UL);
 }
 
-static void pnv_npu2_mn_change_pte(struct mmu_notifier *mn,
-				struct mm_struct *mm,
-				unsigned long address,
-				pte_t pte)
-{
-	struct npu_context *npu_context = mn_to_npu_context(mn);
-	mmio_invalidate(npu_context, address, PAGE_SIZE);
-}
-
 static void pnv_npu2_mn_invalidate_range(struct mmu_notifier *mn,
 					struct mm_struct *mm,
 					unsigned long start, unsigned long end)
@@ -936,7 +927,6 @@ static void pnv_npu2_mn_invalidate_range(struct mmu_notifier *mn,
 
 static const struct mmu_notifier_ops nv_nmmu_notifier_ops = {
 	.release = pnv_npu2_mn_release,
-	.change_pte = pnv_npu2_mn_change_pte,
 	.invalidate_range = pnv_npu2_mn_invalidate_range,
 };
 
Signature · 1 line
2.17.1
Reviewed-by
#2 lore raw
Quoted 8 lines · AEDT Andrea Arcangeli
Quoted 7 lines · Peter Xu

The change_pte() notifier was designed to use as a quick path to update secondary MMU PTEs on write permission changes or PFN changes. For KVM, it could reduce the vm-exits when vcpu faults on the pages that was touched up by KSM. It's not used to do cache invalidations, for example, if we see the notifier will be called before the real PTE update after all (please see set_pte_at_notify that set_pte_at was called later).

Thanks for the fixup. I didn't realise that invalidate_range() always gets called but I now see that is the case so this change looks good to me as well.

Quoted 93 lines
Quoted 17 lines

All the necessary cache invalidation should all be done in invalidate_range() already.

CC Benjamin Herrenschmidt <[email protected]>
CC Paul Mackerras <[email protected]>
CC Michael Ellerman <[email protected]>
CC Alistair Popple <[email protected]>
CC Alexey Kardashevskiy <[email protected]>
CC Mark Hairgrove <[email protected]>
CC Balbir Singh <[email protected]>
CC David Gibson <[email protected]>
CC Andrea Arcangeli <[email protected]>
CC Jerome Glisse <[email protected]>
CC Jason Wang <[email protected]>
Quoted 4 lines

---

 arch/powerpc/platforms/powernv/npu-dma.c | 10 ----------
 1 file changed, 10 deletions(-)

It doesn't make sense to implement change_pte as an invalidate, change_pte is not compulsory to implement so if one wants to have invalidates only, change_pte method shouldn't be implemented in the first place and the common code will guarantee to invoke the range invalidates instead.

Currently the whole change_pte optimization is effectively disabled as noted in past discussions with Jerome (because of the range invalidates that always surrounds it), so we need to revisit the whole change_pte logic and decide it to re-enable it or to drop it as a whole, but in the meantime it's good to cleanup spots like below that should leave change_pte alone.

There are several examples of mmu_notifiers_ops in the kernel that don't implement change_pte, in fact it's the majority. Of all mmu notifier users, only nv_nmmu_notifier_ops, intel_mmuops_change and kvm_mmu_notifier_ops implements change_pte and as Peter found out by source review nv_nmmu_notifier_ops, intel_mmuops_change are wrong about it and should stop implementing it as an invalidate.

In short change_pte is only implemented correctly from KVM which can really updates the spte and flushes the TLB but the spte update remains and could avoid a vmexit if we figure out how to re-enable the optimization safely (the TLB fill after change_pte in KVM EPT/shadow secondary MMU will be looked up by the CPU in hardware).

If change_pte is implemented, it should update the mapping like KVM does and not do an invalidate.

Thanks,
Andrea

Quoted 37 lines
(quoted diff) +0 -0

b/arch/powerpc/platforms/powernv/npu-dma.c index
3f58c7dbd581..c003b29d870e 100644

arch/powerpc/platforms/powernv/npu-dma.c +0 -9
@@ -917,15 +917,6 @@ static void pnv_npu2_mn_release(struct mmu_notifier
 *mn,> 
 	mmio_invalidate(npu_context, 0, ~0UL);
 
 }
 
-static void pnv_npu2_mn_change_pte(struct mmu_notifier *mn,
-				struct mm_struct *mm,
-				unsigned long address,
-				pte_t pte)
-{
-	struct npu_context *npu_context = mn_to_npu_context(mn);
-	mmio_invalidate(npu_context, address, PAGE_SIZE);
-}
-
 

static void pnv_npu2_mn_invalidate_range(struct mmu_notifier *mn,

arch/powerpc/platforms/powernv/npu-dma.c +0 -0
 					struct mm_struct *mm,
 					unsigned long start, unsigned long end)
@@ -936,7 +927,6 @@ static void pnv_npu2_mn_invalidate_range(struct

mmu_notifier *mn,>
static const struct mmu_notifier_ops nv_nmmu_notifier_ops = {

arch/powerpc/platforms/powernv/npu-dma.c +0 -1
 	.release = pnv_npu2_mn_release,
 
-	.change_pte = pnv_npu2_mn_change_pte,
 
 	.invalidate_range = pnv_npu2_mn_invalidate_range,
 
 };
Reviewed-by
#3 lore raw
Quoted 11 lines · Alistair Popple
Quoted 8 lines · AEDT Andrea Arcangeli
Quoted 7 lines · Peter Xu

The change_pte() notifier was designed to use as a quick path to update secondary MMU PTEs on write permission changes or PFN changes. For KVM, it could reduce the vm-exits when vcpu faults on the pages that was touched up by KSM. It's not used to do cache invalidations, for example, if we see the notifier will be called before the real PTE update after all (please see set_pte_at_notify that set_pte_at was called later).

Quoted 5 lines

Thanks for the fixup. I didn't realise that invalidate_range() always gets called but I now see that is the case so this change looks good to me as well.

I checked the three callers of set_pte_at_notify and the assumption seems correct

#4 lore raw
In reply to Peter Xu
Quoted 20 lines · Peter Xu

The change_pte() notifier was designed to use as a quick path to update secondary MMU PTEs on write permission changes or PFN changes. For KVM, it could reduce the vm-exits when vcpu faults on the pages that was touched up by KSM. It's not used to do cache invalidations, for example, if we see the notifier will be called before the real PTE update after all (please see set_pte_at_notify that set_pte_at was called later).

All the necessary cache invalidation should all be done in invalidate_range() already.

CC Benjamin Herrenschmidt <[email protected]>
CC Paul Mackerras <[email protected]>
CC Michael Ellerman <[email protected]>
CC Alistair Popple <[email protected]>
CC Alexey Kardashevskiy <[email protected]>
CC Mark Hairgrove <[email protected]>
CC Balbir Singh <[email protected]>
CC David Gibson <[email protected]>
CC Andrea Arcangeli <[email protected]>
Quoted 8 lines

Applied to powerpc next, thanks.

https://git.kernel.org/powerpc/c/1b58a975be36994a572ae3b3fb3e0232

cheers