[PATCH v3 4/4] iommu/iova: Make dma_32bit_pfn implicit

From: Robin Murphy
Date: Tue Aug 22 2017 - 11:18:13 EST

Next message: Robin Murphy: "[PATCH v3 2/4] iommu/iova: Optimise the padding calculation"
Previous message: Robin Murphy: "[PATCH v3 3/4] iommu/iova: Extend rbtree node caching"
In reply to: Robin Murphy: "[PATCH v3 3/4] iommu/iova: Extend rbtree node caching"
Next in thread: Robin Murphy: "[PATCH v3 2/4] iommu/iova: Optimise the padding calculation"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

From: Zhen Lei <thunder.leizhen@xxxxxxxxxx>

Now that the cached node optimisation can apply to all allocations, the
couple of users which were playing tricks with dma_32bit_pfn in order to
benefit from it can stop doing so. Conversely, there is also no need for
all the other users to explicitly calculate a 'real' 32-bit PFN, when
init_iova_domain() can happily do that itself from the page granularity.

CC: Thierry Reding <thierry.reding@xxxxxxxxx>
CC: Jonathan Hunter <jonathanh@xxxxxxxxxx>
CC: David Airlie <airlied@xxxxxxxx>
CC: Sudeep Dutt <sudeep.dutt@xxxxxxxxx>
CC: Ashutosh Dixit <ashutosh.dixit@xxxxxxxxx>
Signed-off-by: Zhen Lei <thunder.leizhen@xxxxxxxxxx>
Tested-by: Ard Biesheuvel <ard.biesheuvel@xxxxxxxxxx>
Tested-by: Zhen Lei <thunder.leizhen@xxxxxxxxxx>
[rm: use iova_shift(), rewrote commit message]
Signed-off-by: Robin Murphy <robin.murphy@xxxxxxx>
---

v3: No change

drivers/gpu/drm/tegra/drm.c | 3 +--
drivers/gpu/host1x/dev.c | 3 +--
drivers/iommu/amd_iommu.c | 7 ++-----
drivers/iommu/dma-iommu.c | 18 +-----------------
drivers/iommu/intel-iommu.c | 11 +++--------
drivers/iommu/iova.c | 4 ++--
drivers/misc/mic/scif/scif_rma.c | 3 +--
include/linux/iova.h | 5 ++---
8 files changed, 13 insertions(+), 41 deletions(-)

diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c
index 518f4b69ea53..81e9ae1ee90b 100644
--- a/drivers/gpu/drm/tegra/drm.c
+++ b/drivers/gpu/drm/tegra/drm.c
@@ -150,8 +150,7 @@ static int tegra_drm_load(struct drm_device *drm, unsigned long flags)

order = __ffs(tegra->domain->pgsize_bitmap);
init_iova_domain(&tegra->carveout.domain, 1UL << order,
- carveout_start >> order,
- carveout_end >> order);
+ carveout_start >> order);

tegra->carveout.shift = iova_shift(&tegra->carveout.domain);
tegra->carveout.limit = carveout_end >> tegra->carveout.shift;
diff --git a/drivers/gpu/host1x/dev.c b/drivers/gpu/host1x/dev.c
index 2c58a390123a..57c8eed0ed71 100644
--- a/drivers/gpu/host1x/dev.c
+++ b/drivers/gpu/host1x/dev.c
@@ -193,8 +193,7 @@ static int host1x_probe(struct platform_device *pdev)

order = __ffs(host->domain->pgsize_bitmap);
init_iova_domain(&host->iova, 1UL << order,
- geometry->aperture_start >> order,
- geometry->aperture_end >> order);
+ geometry->aperture_start >> order);
host->iova_end = geometry->aperture_end;
}

diff --git a/drivers/iommu/amd_iommu.c b/drivers/iommu/amd_iommu.c
index 688e77576e5a..a12e3e12014a 100644
--- a/drivers/iommu/amd_iommu.c
+++ b/drivers/iommu/amd_iommu.c
@@ -63,7 +63,6 @@
/* IO virtual address start page frame number */
#define IOVA_START_PFN (1)
#define IOVA_PFN(addr) ((addr) >> PAGE_SHIFT)
-#define DMA_32BIT_PFN IOVA_PFN(DMA_BIT_MASK(32))

/* Reserved IOVA ranges */
#define MSI_RANGE_START (0xfee00000)
@@ -2010,8 +2009,7 @@ static struct dma_ops_domain *dma_ops_domain_alloc(void)
if (!dma_dom->domain.pt_root)
goto free_dma_dom;

- init_iova_domain(&dma_dom->iovad, PAGE_SIZE,
- IOVA_START_PFN, DMA_32BIT_PFN);
+ init_iova_domain(&dma_dom->iovad, PAGE_SIZE, IOVA_START_PFN);

/* Initialize reserved ranges */
copy_reserved_iova(&reserved_iova_ranges, &dma_dom->iovad);
@@ -2912,8 +2910,7 @@ static int init_reserved_iova_ranges(void)
struct pci_dev *pdev = NULL;
struct iova *val;

- init_iova_domain(&reserved_iova_ranges, PAGE_SIZE,
- IOVA_START_PFN, DMA_32BIT_PFN);
+ init_iova_domain(&reserved_iova_ranges, PAGE_SIZE, IOVA_START_PFN);

lockdep_set_class(&reserved_iova_ranges.iova_rbtree_lock,
&reserved_rbtree_key);
diff --git a/drivers/iommu/dma-iommu.c b/drivers/iommu/dma-iommu.c
index 9d1cebe7f6cb..191be9c80a8a 100644
--- a/drivers/iommu/dma-iommu.c
+++ b/drivers/iommu/dma-iommu.c
@@ -292,18 +292,7 @@ int iommu_dma_init_domain(struct iommu_domain *domain, dma_addr_t base,
/* ...then finally give it a kicking to make sure it fits */
base_pfn = max_t(unsigned long, base_pfn,
domain->geometry.aperture_start >> order);
- end_pfn = min_t(unsigned long, end_pfn,
- domain->geometry.aperture_end >> order);
}
- /*
- * PCI devices may have larger DMA masks, but still prefer allocating
- * within a 32-bit mask to avoid DAC addressing. Such limitations don't
- * apply to the typical platform device, so for those we may as well
- * leave the cache limit at the top of their range to save an rb_last()
- * traversal on every allocation.
- */
- if (dev && dev_is_pci(dev))
- end_pfn &= DMA_BIT_MASK(32) >> order;

/* start_pfn is always nonzero for an already-initialised domain */
if (iovad->start_pfn) {
@@ -312,16 +301,11 @@ int iommu_dma_init_domain(struct iommu_domain *domain, dma_addr_t base,
pr_warn("Incompatible range for DMA domain\n");
return -EFAULT;
}
- /*
- * If we have devices with different DMA masks, move the free
- * area cache limit down for the benefit of the smaller one.
- */
- iovad->dma_32bit_pfn = min(end_pfn + 1, iovad->dma_32bit_pfn);

return 0;
}

- init_iova_domain(iovad, 1UL << order, base_pfn, end_pfn);
+ init_iova_domain(iovad, 1UL << order, base_pfn);
if (!dev)
return 0;

diff --git a/drivers/iommu/intel-iommu.c b/drivers/iommu/intel-iommu.c
index 687f18f65cea..afa3b4e765e7 100644
--- a/drivers/iommu/intel-iommu.c
+++ b/drivers/iommu/intel-iommu.c
@@ -82,8 +82,6 @@
#define IOVA_START_PFN (1)

#define IOVA_PFN(addr) ((addr) >> PAGE_SHIFT)
-#define DMA_32BIT_PFN IOVA_PFN(DMA_BIT_MASK(32))
-#define DMA_64BIT_PFN IOVA_PFN(DMA_BIT_MASK(64))

/* page table handling */
#define LEVEL_STRIDE (9)
@@ -1874,8 +1872,7 @@ static int dmar_init_reserved_ranges(void)
struct iova *iova;
int i;

- init_iova_domain(&reserved_iova_list, VTD_PAGE_SIZE, IOVA_START_PFN,
- DMA_32BIT_PFN);
+ init_iova_domain(&reserved_iova_list, VTD_PAGE_SIZE, IOVA_START_PFN);

lockdep_set_class(&reserved_iova_list.iova_rbtree_lock,
&reserved_rbtree_key);
@@ -1933,8 +1930,7 @@ static int domain_init(struct dmar_domain *domain, struct intel_iommu *iommu,
int adjust_width, agaw;
unsigned long sagaw;

- init_iova_domain(&domain->iovad, VTD_PAGE_SIZE, IOVA_START_PFN,
- DMA_32BIT_PFN);
+ init_iova_domain(&domain->iovad, VTD_PAGE_SIZE, IOVA_START_PFN);
domain_reserve_special_ranges(domain);

/* calculate AGAW */
@@ -4989,8 +4985,7 @@ static int md_domain_init(struct dmar_domain *domain, int guest_width)
{
int adjust_width;

- init_iova_domain(&domain->iovad, VTD_PAGE_SIZE, IOVA_START_PFN,
- DMA_32BIT_PFN);
+ init_iova_domain(&domain->iovad, VTD_PAGE_SIZE, IOVA_START_PFN);
domain_reserve_special_ranges(domain);

/* calculate AGAW */
diff --git a/drivers/iommu/iova.c b/drivers/iommu/iova.c
index b208f656f6a4..612b8f6ce34d 100644
--- a/drivers/iommu/iova.c
+++ b/drivers/iommu/iova.c
@@ -35,7 +35,7 @@ static void free_iova_rcaches(struct iova_domain *iovad);

void
init_iova_domain(struct iova_domain *iovad, unsigned long granule,
- unsigned long start_pfn, unsigned long pfn_32bit)
+ unsigned long start_pfn)
{
/*
* IOVA granularity will normally be equal to the smallest
@@ -50,7 +50,7 @@ init_iova_domain(struct iova_domain *iovad, unsigned long granule,
iovad->cached32_node = NULL;
iovad->granule = granule;
iovad->start_pfn = start_pfn;
- iovad->dma_32bit_pfn = pfn_32bit + 1;
+ iovad->dma_32bit_pfn = 1UL << (32 - iova_shift(iovad));
init_iova_rcaches(iovad);
}
EXPORT_SYMBOL_GPL(init_iova_domain);
diff --git a/drivers/misc/mic/scif/scif_rma.c b/drivers/misc/mic/scif/scif_rma.c
index 329727e00e97..c824329f7012 100644
--- a/drivers/misc/mic/scif/scif_rma.c
+++ b/drivers/misc/mic/scif/scif_rma.c
@@ -39,8 +39,7 @@ void scif_rma_ep_init(struct scif_endpt *ep)
struct scif_endpt_rma_info *rma = &ep->rma_info;

mutex_init(&rma->rma_lock);
- init_iova_domain(&rma->iovad, PAGE_SIZE, SCIF_IOVA_START_PFN,
- SCIF_DMA_64BIT_PFN);
+ init_iova_domain(&rma->iovad, PAGE_SIZE, SCIF_IOVA_START_PFN);
spin_lock_init(&rma->tc_lock);
mutex_init(&rma->mmn_lock);
INIT_LIST_HEAD(&rma->reg_list);
diff --git a/include/linux/iova.h b/include/linux/iova.h
index 0bb8df43b393..58c2a365c45f 100644
--- a/include/linux/iova.h
+++ b/include/linux/iova.h
@@ -102,7 +102,7 @@ struct iova *reserve_iova(struct iova_domain *iovad, unsigned long pfn_lo,
unsigned long pfn_hi);
void copy_reserved_iova(struct iova_domain *from, struct iova_domain *to);
void init_iova_domain(struct iova_domain *iovad, unsigned long granule,
- unsigned long start_pfn, unsigned long pfn_32bit);
+ unsigned long start_pfn);
struct iova *find_iova(struct iova_domain *iovad, unsigned long pfn);
void put_iova_domain(struct iova_domain *iovad);
struct iova *split_and_remove_iova(struct iova_domain *iovad,
@@ -170,8 +170,7 @@ static inline void copy_reserved_iova(struct iova_domain *from,

static inline void init_iova_domain(struct iova_domain *iovad,
unsigned long granule,
- unsigned long start_pfn,
- unsigned long pfn_32bit)
+ unsigned long start_pfn)
{
}

--
2.13.4.dirty

Next message: Robin Murphy: "[PATCH v3 2/4] iommu/iova: Optimise the padding calculation"
Previous message: Robin Murphy: "[PATCH v3 3/4] iommu/iova: Extend rbtree node caching"
In reply to: Robin Murphy: "[PATCH v3 3/4] iommu/iova: Extend rbtree node caching"
Next in thread: Robin Murphy: "[PATCH v3 2/4] iommu/iova: Optimise the padding calculation"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]