All of lore.kernel.org
 help / color / mirror / Atom feed
From: Will Deacon <will@kernel.org>
To: iommu@lists.linux-foundation.org
Cc: "Isaac J. Manjarres" <isaacm@codeaurora.org>,
	Will Deacon <will@kernel.org>,
	Robin Murphy <robin.murphy@arm.com>,
	Pratik Patel <pratikp@codeaurora.org>,
	linux-arm-kernel@lists.infradead.org
Subject: [RFC PATCH 6/6] iommu: Accomodate larger pages in iommu_pgsize() 'count' calculation
Date: Thu,  1 Apr 2021 17:47:38 +0100	[thread overview]
Message-ID: <20210401164738.9513-7-will@kernel.org> (raw)
In-Reply-To: <20210401164738.9513-1-will@kernel.org>

Extend the calculation of 'count' in iommu_pgsize() so that it takes
larger page sizes into consideration and returns a value which will
allow a larger page size to be used on the next call.

Signed-off-by: Will Deacon <will@kernel.org>
---
 drivers/iommu/iommu.c | 36 ++++++++++++++++++++++++++++++++----
 1 file changed, 32 insertions(+), 4 deletions(-)

diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index fe186691fc21..0572a4dcd9e2 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c
@@ -2360,9 +2360,9 @@ EXPORT_SYMBOL_GPL(iommu_iova_to_phys);
 static size_t iommu_pgsize(struct iommu_domain *domain, unsigned long iova,
 			   phys_addr_t paddr, size_t size, size_t *count)
 {
-	unsigned int pgsize_idx;
+	unsigned int pgsize_idx, pgsize_idx_next;
 	unsigned long pgsizes;
-	size_t pgsize;
+	size_t offset, pgsize, pgsize_next;
 	phys_addr_t addr_merge = paddr | iova;
 
 	/* Page sizes supported by the hardware and small enough for @size */
@@ -2378,9 +2378,37 @@ static size_t iommu_pgsize(struct iommu_domain *domain, unsigned long iova,
 	/* Pick the biggest page size remaining */
 	pgsize_idx = __fls(pgsizes);
 	pgsize = BIT(pgsize_idx);
+	if (!count)
+		return pgsize;
 
-	if (count)
-		*count = size >> pgsize_idx;
+
+	/* Find the next biggest support page size, if it exists */
+	pgsizes = domain->pgsize_bitmap & ~GENMASK(pgsize_idx, 0);
+	if (!pgsizes)
+		goto out_set_count;
+
+	pgsize_idx_next = __ffs(pgsizes);
+	pgsize_next = BIT(pgsize_idx_next);
+
+	/*
+	 * There's no point trying a bigger page size unless the virtual
+	 * and physical addresses are similarly offset within the larger page.
+	 */
+	if ((iova ^ paddr) & (pgsize_next - 1))
+		goto out_set_count;
+
+	/* Calculate the offset to the next page size alignment boundary */
+	offset = pgsize_next - (addr_merge & (pgsize_next - 1));
+
+	/*
+	 * If size is big enough to accomodate the larger page, reduce
+	 * the number of smaller pages.
+	 */
+	if (offset + pgsize_next <= size)
+		size = offset;
+
+out_set_count:
+	*count = size >> pgsize_idx;
 	return pgsize;
 }
 
-- 
2.31.0.291.g576ba9dcdaf-goog

_______________________________________________
iommu mailing list
iommu@lists.linux-foundation.org
https://lists.linuxfoundation.org/mailman/listinfo/iommu

WARNING: multiple messages have this Message-ID (diff)
From: Will Deacon <will@kernel.org>
To: iommu@lists.linux-foundation.org
Cc: linux-arm-kernel@lists.infradead.org,
	Will Deacon <will@kernel.org>,
	"Isaac J. Manjarres" <isaacm@codeaurora.org>,
	Pratik Patel <pratikp@codeaurora.org>,
	Robin Murphy <robin.murphy@arm.com>,
	Lu Baolu <baolu.lu@linux.intel.com>
Subject: [RFC PATCH 6/6] iommu: Accomodate larger pages in iommu_pgsize() 'count' calculation
Date: Thu,  1 Apr 2021 17:47:38 +0100	[thread overview]
Message-ID: <20210401164738.9513-7-will@kernel.org> (raw)
In-Reply-To: <20210401164738.9513-1-will@kernel.org>

Extend the calculation of 'count' in iommu_pgsize() so that it takes
larger page sizes into consideration and returns a value which will
allow a larger page size to be used on the next call.

Signed-off-by: Will Deacon <will@kernel.org>
---
 drivers/iommu/iommu.c | 36 ++++++++++++++++++++++++++++++++----
 1 file changed, 32 insertions(+), 4 deletions(-)

diff --git a/drivers/iommu/iommu.c b/drivers/iommu/iommu.c
index fe186691fc21..0572a4dcd9e2 100644
--- a/drivers/iommu/iommu.c
+++ b/drivers/iommu/iommu.c
@@ -2360,9 +2360,9 @@ EXPORT_SYMBOL_GPL(iommu_iova_to_phys);
 static size_t iommu_pgsize(struct iommu_domain *domain, unsigned long iova,
 			   phys_addr_t paddr, size_t size, size_t *count)
 {
-	unsigned int pgsize_idx;
+	unsigned int pgsize_idx, pgsize_idx_next;
 	unsigned long pgsizes;
-	size_t pgsize;
+	size_t offset, pgsize, pgsize_next;
 	phys_addr_t addr_merge = paddr | iova;
 
 	/* Page sizes supported by the hardware and small enough for @size */
@@ -2378,9 +2378,37 @@ static size_t iommu_pgsize(struct iommu_domain *domain, unsigned long iova,
 	/* Pick the biggest page size remaining */
 	pgsize_idx = __fls(pgsizes);
 	pgsize = BIT(pgsize_idx);
+	if (!count)
+		return pgsize;
 
-	if (count)
-		*count = size >> pgsize_idx;
+
+	/* Find the next biggest support page size, if it exists */
+	pgsizes = domain->pgsize_bitmap & ~GENMASK(pgsize_idx, 0);
+	if (!pgsizes)
+		goto out_set_count;
+
+	pgsize_idx_next = __ffs(pgsizes);
+	pgsize_next = BIT(pgsize_idx_next);
+
+	/*
+	 * There's no point trying a bigger page size unless the virtual
+	 * and physical addresses are similarly offset within the larger page.
+	 */
+	if ((iova ^ paddr) & (pgsize_next - 1))
+		goto out_set_count;
+
+	/* Calculate the offset to the next page size alignment boundary */
+	offset = pgsize_next - (addr_merge & (pgsize_next - 1));
+
+	/*
+	 * If size is big enough to accomodate the larger page, reduce
+	 * the number of smaller pages.
+	 */
+	if (offset + pgsize_next <= size)
+		size = offset;
+
+out_set_count:
+	*count = size >> pgsize_idx;
 	return pgsize;
 }
 
-- 
2.31.0.291.g576ba9dcdaf-goog


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

  parent reply	other threads:[~2021-04-01 16:48 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-01 16:47 [RFC PATCH 0/6] iommu_pgsize() improvements to help towards ->[un]map_pages() Will Deacon
2021-04-01 16:47 ` Will Deacon
2021-04-01 16:47 ` [RFC PATCH 1/6] iommu/io-pgtable: Introduce unmap_pages() as a page table op Will Deacon
2021-04-01 16:47   ` Will Deacon
2021-04-01 16:47 ` [RFC PATCH 2/6] iommu: Add an unmap_pages() op for IOMMU drivers Will Deacon
2021-04-01 16:47   ` Will Deacon
2021-04-01 16:47 ` [RFC PATCH 3/6] iommu: Use bitmap to calculate page size in iommu_pgsize() Will Deacon
2021-04-01 16:47   ` Will Deacon
2021-04-02  1:39   ` isaacm
2021-04-06 11:40     ` Will Deacon
2021-04-06 11:40       ` Will Deacon
2021-04-01 16:47 ` [RFC PATCH 4/6] iommu: Split 'addr_merge' argument to iommu_pgsize() into separate parts Will Deacon
2021-04-01 16:47   ` Will Deacon
2021-04-01 16:47 ` [RFC PATCH 5/6] iommu: Hook up '->unmap_pages' driver callback Will Deacon
2021-04-01 16:47   ` Will Deacon
2021-04-01 16:47 ` Will Deacon [this message]
2021-04-01 16:47   ` [RFC PATCH 6/6] iommu: Accomodate larger pages in iommu_pgsize() 'count' calculation Will Deacon

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20210401164738.9513-7-will@kernel.org \
    --to=will@kernel.org \
    --cc=iommu@lists.linux-foundation.org \
    --cc=isaacm@codeaurora.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=pratikp@codeaurora.org \
    --cc=robin.murphy@arm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body. 
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.