iommu/io-pgtable-arm: Optimise non-coherent unmap
Subsystems (3)
PATCH v4 1/1 iommu/io-pgtable-arm: Optimise non-coherent unmap
The current __arm_lpae_unmap() function calls dma_sync() on individual PTEs after clearing them. Overall unmap performance can be improved by around 25% for large buffer sizes by combining the syncs for adjacent leaf entries.
Optimize the unmap time by clearing all the leaf entries and issuing a single dma_sync() for them.
Below is detailed analysis of average unmap latency(in us) with and without this optimization obtained by running dma_map_benchmark for different buffer sizes.
UnMap Latency(us)
Size Without With % gain with
optimiztion optimization optimization
4KB 3 3 0
8KB 4 3.8 5
16KB 6.1 5.4 11.48
32KB 10.2 8.5 16.67
64KB 18.5 14.9 19.46
128KB 35 27.5 21.43
256KB 67.5 52.2 22.67
512KB 127.9 97.2 24.00
1MB 248.6 187.4 24.62
2MB 65.5 65.5 0
4MB 119.2 119 0.17
---
Changes in V2:
- Updated the commit message to be imperative.
- Fixed ptep at incorrect index getting cleared for non-leaf entries.
Changes in V3:
- Used loop-local variables and removed redundant function variables.
- Added check for zero-sized dma_sync in __arm_lpae_clear_pte().
- Merged both patches into this single patch by adding check for a NULL gather in __arm_lpae_unmap() itself.
Changes in V4:
- Updated the subject in commit message to correctly reflect the changes made in this patch.
--- drivers/iommu/io-pgtable-arm.c | 31 +++++++++++++++++-------------- 1 file changed, 17 insertions(+), 14 deletions(-)
drivers/iommu/io-pgtable-arm.c +17 -14
@@ -274,13 +274,13 @@ static void __arm_lpae_sync_pte(arm_lpae_iopte *ptep, int num_entries, sizeof(*ptep) * num_entries, DMA_TO_DEVICE); } -static void __arm_lpae_clear_pte (arm_lpae_iopte *ptep,struct io_pgtable_cfg *cfg) +static void __arm_lpae_clear_pte (arm_lpae_iopte *ptep,struct io_pgtable_cfg *cfg,int num_entries) { +for (int i =0 ; i < num_entries; i++) + ptep[i] =0 ; - *ptep =0 ; - -if (!cfg->coherent_walk) -__arm_lpae_sync_pte (ptep,1 , cfg); +if (!cfg->coherent_walk && num_entries) +__arm_lpae_sync_pte (ptep, num_entries, cfg); } static size_t __arm_lpae_unmap(struct arm_lpae_io_pgtable *data,
@@ -654,26 +654,29 @@ static size_t __arm_lpae_unmap(struct arm_lpae_io_pgtable *data, max_entries =ARM_LPAE_PTES_PER_TABLE (data) - unmap_idx_start; num_entries =min_t (int, pgcount, max_entries); -while (i < num_entries) { - pte =READ_ONCE (*ptep); +/* Find and handle non-leaf entries */ +for (i =0 ; i < num_entries; i++) { + pte =READ_ONCE (ptep[i]);if (WARN_ON (!pte))break ; -__arm_lpae_clear_pte (ptep, &iop->cfg); -if (!iopte_leaf (pte, lvl, iop->fmt)) { +__arm_lpae_clear_pte (&ptep[i], &iop->cfg,1 ); +/* Also flush any partial walks */ io_pgtable_tlb_flush_walk (iop, iova + i * size, size,ARM_LPAE_GRANULE (data));__arm_lpae_free_pgtable (data, lvl +1 ,iopte_deref (pte, data)); - }else if (!iommu_iotlb_gather_queued (gather)) { -io_pgtable_tlb_add_page (iop, gather, iova + i * size, size); } - - ptep++; - i++; } +/* Clear the remaining entries */ +__arm_lpae_clear_pte (ptep, &iop->cfg, i); + +if (gather && !iommu_iotlb_gather_queued (gather)) +for (int j =0 ; j < i; j++) +io_pgtable_tlb_add_page (iop, gather, iova + j * size, size); +return i * size; }else if (iopte_leaf (pte, lvl, iop->fmt)) {/*
Signature · 1 line
2.25.1
Quoted 98 lines · Ashish Mhetre
The current
__arm_lpae_unmap()function callsdma_sync()on individual PTEs after clearing them. Overall unmap performance can be improved by around 25% for large buffer sizes by combining the syncs for adjacent leaf entries.
Optimize the unmap time by clearing all the leaf entries and issuing a singledma_sync()for them.
Below is detailed analysis of average unmap latency(in us) with and without this optimization obtained by runningdma_map_benchmarkfor different buffer sizes.UnMap Latency(us)
Size Without With % gain with
optimiztion optimization optimization4KB 3 3 0
8KB 4 3.8 5
16KB 6.1 5.4 11.48
32KB 10.2 8.5 16.67
64KB 18.5 14.9 19.46
128KB 35 27.5 21.43
256KB 67.5 52.2 22.67
512KB 127.9 97.2 24.00
1MB 248.6 187.4 24.62
2MB 65.5 65.5 0
4MB 119.2 119 0.17Reviewed-by Robin Murphy <[email protected]>Signed-off-by Ashish Mhetre <[email protected]>---
Changes in V2:
- Updated the commit message to be imperative.
- Fixed ptep at incorrect index getting cleared for non-leaf entries.Changes in V3:
- Used loop-local variables and removed redundant function variables.
- Added check for zero-sizeddma_syncin__arm_lpae_clear_pte().
- Merged both patches into this single patch by adding check for a NULL gather in__arm_lpae_unmap()itself.Changes in V4:
- Updated the subject in commit message to correctly reflect the changes made in this patch.--- drivers/iommu/io-pgtable-arm.c | 31 +++++++++++++++++-------------- 1 file changed, 17 insertions(+), 14 deletions(-)drivers/iommu/io-pgtable-arm.c +10
-14@@ -274,13 +274,13 @@ static void __arm_lpae_sync_pte(arm_lpae_iopte *ptep, int num_entries, sizeof(*ptep) * num_entries, DMA_TO_DEVICE); } -static void __arm_lpae_clear_pte (arm_lpae_iopte *ptep,struct io_pgtable_cfg *cfg) +static void __arm_lpae_clear_pte (arm_lpae_iopte *ptep,struct io_pgtable_cfg *cfg,int num_entries) { +for (int i =0 ; i < num_entries; i++) + ptep[i] =0 ; - *ptep =0 ; - -if (!cfg->coherent_walk) -__arm_lpae_sync_pte (ptep,1 , cfg); +if (!cfg->coherent_walk && num_entries) +__arm_lpae_sync_pte (ptep, num_entries, cfg); } static size_t __arm_lpae_unmap(struct arm_lpae_io_pgtable *data,@@ -654,26 +654,29 @@ static size_t __arm_lpae_unmap(struct arm_lpae_io_pgtable *data, max_entries =ARM_LPAE_PTES_PER_TABLE (data) - unmap_idx_start; num_entries =min_t (int, pgcount, max_entries); -while (i < num_entries) { - pte =READ_ONCE (*ptep); +/* Find and handle non-leaf entries */ +for (i =0 ; i < num_entries; i++) { + pte =READ_ONCE (ptep[i]);if (WARN_ON (!pte))break ; -__arm_lpae_clear_pte (ptep, &iop->cfg); -if (!iopte_leaf (pte, lvl, iop->fmt)) { +__arm_lpae_clear_pte (&ptep[i], &iop->cfg,1 ); +/* Also flush any partial walks */ io_pgtable_tlb_flush_walk (iop, iova + i * size, size,ARM_LPAE_GRANULE (data));__arm_lpae_free_pgtable (data, lvl +1 ,iopte_deref (pte, data)); - }else if (!iommu_iotlb_gather_queued (gather)) { -io_pgtable_tlb_add_page (iop, gather, iova + i * size, size); } - - ptep++; - i++; }
drivers/iommu/io-pgtable-arm.c static size_t __arm_lpae_unmap(struct arm_lpae_io_pgtable *data, · line 682
Quoted 10 lines
drivers/iommu/io-pgtable-arm.c +7
-0+/* Clear the remaining entries */ +__arm_lpae_clear_pte (ptep, &iop->cfg, i); + +if (gather && !iommu_iotlb_gather_queued (gather)) +for (int j =0 ; j < i; j++) +io_pgtable_tlb_add_page (iop, gather, iova + j * size, size); +return i * size; }else if (iopte_leaf (pte, lvl, iop->fmt)) {/*
Hi all,
Can you please review this patch and provide feedback?
Thanks,
Ashish Mhetre
Quoted 122 lines · Ashish Mhetre
External email: Use caution opening links or attachments
Quoted 110 lines · Ashish Mhetre
The current
__arm_lpae_unmap()function callsdma_sync()on individual PTEs after clearing them. Overall unmap performance can be improved by around 25% for large buffer sizes by combining the syncs for adjacent leaf entries.
Optimize the unmap time by clearing all the leaf entries and issuing a singledma_sync()for them.
Below is detailed analysis of average unmap latency(in us) with and without this optimization obtained by runningdma_map_benchmarkfor different buffer sizes.UnMap Latency(us)
Size Without With % gain with optimiztion optimization optimization4KB 3 3 0
8KB 4 3.8 5
16KB 6.1 5.4 11.48
32KB 10.2 8.5 16.67
64KB 18.5 14.9 19.46
128KB 35 27.5 21.43
256KB 67.5 52.2 22.67
512KB 127.9 97.2 24.00
1MB 248.6 187.4 24.62
2MB 65.5 65.5 0
4MB 119.2 119 0.17Reviewed-by Robin Murphy <[email protected]>Signed-off-by Ashish Mhetre <[email protected]>---
Changes in V2:
- Updated the commit message to be imperative.
- Fixed ptep at incorrect index getting cleared for non-leaf entries.Changes in V3:
- Used loop-local variables and removed redundant function variables.
- Added check for zero-sizeddma_syncin__arm_lpae_clear_pte().
- Merged both patches into this single patch by adding check for a NULL gather in__arm_lpae_unmap()itself.Changes in V4:
- Updated the subject in commit message to correctly reflect the changes made in this patch.
---
drivers/iommu/io-pgtable-arm.c| 31 +++++++++++++++++--------------1 file changed, 17 insertions(+), 14 deletions(-)drivers/iommu/io-pgtable-arm.c +5
-5@@ -274,13 +274,13 @@ static void __arm_lpae_sync_pte(arm_lpae_iopte *ptep, int num_entries, sizeof(*ptep) * num_entries, DMA_TO_DEVICE); } -static void __arm_lpae_clear_pte(arm_lpae_iopte *ptep, struct io_pgtable_cfg *cfg) +static void __arm_lpae_clear_pte (arm_lpae_iopte *ptep,struct io_pgtable_cfg *cfg,int num_entries) { + for (int i = 0; i < num_entries; i++) + ptep[i] = 0; - *ptep = 0; - - if (!cfg->coherent_walk) - __arm_lpae_sync_pte(ptep, 1, cfg); + if (!cfg->coherent_walk && num_entries) + __arm_lpae_sync_pte(ptep, num_entries, cfg);}
static
size_t__arm_lpae_unmap(structarm_lpae_io_pgtable*data,drivers/iommu/io-pgtable-arm.c +11
-9@@ -654,26 +654,29 @@ static size_t __arm_lpae_unmap(structarm_lpae_io_pgtable *data, max_entries = ARM_LPAE_PTES_PER_TABLE(data) - unmap_idx_start; num_entries = min_t(int, pgcount, max_entries); - while (i < num_entries) { - pte = READ_ONCE(*ptep); + /* Find and handle non-leaf entries */ + for (i = 0; i < num_entries; i++) { + pte = READ_ONCE(ptep[i]); if (WARN_ON(!pte)) break; - __arm_lpae_clear_pte(ptep, &iop->cfg); - if (!iopte_leaf(pte, lvl, iop->fmt)) { + __arm_lpae_clear_pte(&ptep[i], &iop->cfg,1 ); + /* Also flush any partial walks */ io_pgtable_tlb_flush_walk(iop, iova + i * size, size,ARM_LPAE_GRANULE (data)); __arm_lpae_free_pgtable(data, lvl + 1, iopte_deref(pte, data)); - } else if (!iommu_iotlb_gather_queued(gather)) { - io_pgtable_tlb_add_page(iop, gather, iova + i * size, size); } - - ptep++; - i++; } + /* Clear the remaining entries */ + __arm_lpae_clear_pte(ptep, &iop->cfg, i); + + if (gather && !iommu_iotlb_gather_queued(gather)) + for (int j = 0; j < i; j++) + io_pgtable_tlb_add_page(iop, gather,Quoted 5 lines
iova + j * size, size);
+
return i * size;
} else if (iopte_leaf(pte, lvl, iop->fmt)) { /*Hi all,
Quoted 6 lines
Can you please review this patch and provide feedback?
Thanks,
Ashish Mhetre
Hi Will, Joerg,
The patch is reviewed by Robin. If there are no more comments, can you please see if you can merge it?
Thanks and Regards,
Ashish Mhetre
Quoted 39 lines · Ashish Mhetre
The current
__arm_lpae_unmap()function callsdma_sync()on individual PTEs after clearing them. Overall unmap performance can be improved by around 25% for large buffer sizes by combining the syncs for adjacent leaf entries.
Optimize the unmap time by clearing all the leaf entries and issuing a singledma_sync()for them.
Below is detailed analysis of average unmap latency(in us) with and without this optimization obtained by runningdma_map_benchmarkfor different buffer sizes.UnMap Latency(us)
Size Without With % gain with
optimiztion optimization optimization4KB 3 3 0
8KB 4 3.8 5
16KB 6.1 5.4 11.48
32KB 10.2 8.5 16.67
64KB 18.5 14.9 19.46
128KB 35 27.5 21.43
256KB 67.5 52.2 22.67
512KB 127.9 97.2 24.00
1MB 248.6 187.4 24.62
2MB 65.5 65.5 0
4MB 119.2 119 0.17Reviewed-by Robin Murphy <[email protected]>Signed-off-by Ashish Mhetre <[email protected]>---
Changes in V2:
- Updated the commit message to be imperative.
- Fixed ptep at incorrect index getting cleared for non-leaf entries.Changes in V3:
- Used loop-local variables and removed redundant function variables.
- Added check for zero-sizeddma_syncin__arm_lpae_clear_pte().
- Merged both patches into this single patch by adding check for a NULL gather in__arm_lpae_unmap()itself.
Quoted 6 lines
Changes in V4:
- Updated the subject in commit message to correctly reflect the changes made in this patch.--- drivers/iommu/io-pgtable-arm.c | 31 +++++++++++++++++-------------- 1 file changed, 17 insertions(+), 14 deletions(-)
Joerg, please can you pick this one up for -next?
Cheers,
Will
Quoted 2 lines · Ashish Mhetre
drivers/iommu/io-pgtable-arm.c | 31 +++++++++++++++++-------------- 1 file changed, 17 insertions(+), 14 deletions(-)
Applied, thanks.