From: Will Deacon <will.deacon-5wv7dgnIgG8@public.gmane.org> To: Robin Murphy <robin.murphy-5wv7dgnIgG8@public.gmane.org> Cc: iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org, linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org Subject: Re: [PATCH v2] iommu/arm-smmu: Invalidate TLBs properly Date: Mon, 7 Dec 2015 18:28:21 +0000 [thread overview] Message-ID: <20151207182821.GH26191@arm.com> (raw) In-Reply-To: <ac2d6aedf473cf01eb1df48a1f81614f0f74b0b1.1449501523.git.robin.murphy-5wv7dgnIgG8@public.gmane.org> On Mon, Dec 07, 2015 at 06:18:52PM +0000, Robin Murphy wrote: > When invalidating an IOVA range potentially spanning multiple pages, > such as when removing an entire intermediate-level table, we currently > only issue an invalidation for the first IOVA of that range. Since the > architecture specifies that address-based TLB maintenance operations > target a single entry, an SMMU could feasibly retain live entries for > subsequent pages within that unmapped range, which is not good. > > Make sure we hit every possible entry by iterating over the whole range > at the granularity provided by the pagetable implementation. > > Signed-off-by: Robin Murphy <robin.murphy-5wv7dgnIgG8@public.gmane.org> > --- > > v2: include SMMUv3 fix, use the same shorter loop construct everywhere. > > drivers/iommu/arm-smmu-v3.c | 5 ++++- > drivers/iommu/arm-smmu.c | 16 +++++++++++++--- > 2 files changed, 17 insertions(+), 4 deletions(-) > > diff --git a/drivers/iommu/arm-smmu-v3.c b/drivers/iommu/arm-smmu-v3.c > index c302b65..8bb5abf 100644 > --- a/drivers/iommu/arm-smmu-v3.c > +++ b/drivers/iommu/arm-smmu-v3.c > @@ -1354,7 +1354,10 @@ static void arm_smmu_tlb_inv_range_nosync(unsigned long iova, size_t size, > cmd.tlbi.vmid = smmu_domain->s2_cfg.vmid; > } > > - arm_smmu_cmdq_issue_cmd(smmu, &cmd); > + do { > + arm_smmu_cmdq_issue_cmd(smmu, &cmd); > + cmd.tlbi.addr += granule; > + } while (size -= granule); > } > > static struct iommu_gather_ops arm_smmu_gather_ops = { > diff --git a/drivers/iommu/arm-smmu.c b/drivers/iommu/arm-smmu.c > index 601e3dd..eb28c3e 100644 > --- a/drivers/iommu/arm-smmu.c > +++ b/drivers/iommu/arm-smmu.c > @@ -597,12 +597,18 @@ static void arm_smmu_tlb_inv_range_nosync(unsigned long iova, size_t size, > if (!IS_ENABLED(CONFIG_64BIT) || smmu->version == ARM_SMMU_V1) { > iova &= ~12UL; > iova |= ARM_SMMU_CB_ASID(cfg); > - writel_relaxed(iova, reg); > + do { > + writel_relaxed(iova, reg); > + iova += granule; > + } while (size -= granule); > #ifdef CONFIG_64BIT > } else { > iova >>= 12; > iova |= (u64)ARM_SMMU_CB_ASID(cfg) << 48; > - writeq_relaxed(iova, reg); > + do { > + writeq_relaxed(iova, reg); > + iova += granule >> 12; > + } while (size -= granule) This doesn't compile. > #endif > } > #ifdef CONFIG_64BIT > @@ -610,7 +616,11 @@ static void arm_smmu_tlb_inv_range_nosync(unsigned long iova, size_t size, > reg = ARM_SMMU_CB_BASE(smmu) + ARM_SMMU_CB(smmu, cfg->cbndx); > reg += leaf ? ARM_SMMU_CB_S2_TLBIIPAS2L : > ARM_SMMU_CB_S2_TLBIIPAS2; > - writeq_relaxed(iova >> 12, reg); > + iova >>= 12; > + do { > + writeq_relaxed(iova, reg); > + iova += granule >> 12; > + } while (size -= granule) Same here. Please at least build your patches, and preferably test them too. Will
WARNING: multiple messages have this Message-ID (diff)
From: will.deacon@arm.com (Will Deacon) To: linux-arm-kernel@lists.infradead.org Subject: [PATCH v2] iommu/arm-smmu: Invalidate TLBs properly Date: Mon, 7 Dec 2015 18:28:21 +0000 [thread overview] Message-ID: <20151207182821.GH26191@arm.com> (raw) In-Reply-To: <ac2d6aedf473cf01eb1df48a1f81614f0f74b0b1.1449501523.git.robin.murphy@arm.com> On Mon, Dec 07, 2015 at 06:18:52PM +0000, Robin Murphy wrote: > When invalidating an IOVA range potentially spanning multiple pages, > such as when removing an entire intermediate-level table, we currently > only issue an invalidation for the first IOVA of that range. Since the > architecture specifies that address-based TLB maintenance operations > target a single entry, an SMMU could feasibly retain live entries for > subsequent pages within that unmapped range, which is not good. > > Make sure we hit every possible entry by iterating over the whole range > at the granularity provided by the pagetable implementation. > > Signed-off-by: Robin Murphy <robin.murphy@arm.com> > --- > > v2: include SMMUv3 fix, use the same shorter loop construct everywhere. > > drivers/iommu/arm-smmu-v3.c | 5 ++++- > drivers/iommu/arm-smmu.c | 16 +++++++++++++--- > 2 files changed, 17 insertions(+), 4 deletions(-) > > diff --git a/drivers/iommu/arm-smmu-v3.c b/drivers/iommu/arm-smmu-v3.c > index c302b65..8bb5abf 100644 > --- a/drivers/iommu/arm-smmu-v3.c > +++ b/drivers/iommu/arm-smmu-v3.c > @@ -1354,7 +1354,10 @@ static void arm_smmu_tlb_inv_range_nosync(unsigned long iova, size_t size, > cmd.tlbi.vmid = smmu_domain->s2_cfg.vmid; > } > > - arm_smmu_cmdq_issue_cmd(smmu, &cmd); > + do { > + arm_smmu_cmdq_issue_cmd(smmu, &cmd); > + cmd.tlbi.addr += granule; > + } while (size -= granule); > } > > static struct iommu_gather_ops arm_smmu_gather_ops = { > diff --git a/drivers/iommu/arm-smmu.c b/drivers/iommu/arm-smmu.c > index 601e3dd..eb28c3e 100644 > --- a/drivers/iommu/arm-smmu.c > +++ b/drivers/iommu/arm-smmu.c > @@ -597,12 +597,18 @@ static void arm_smmu_tlb_inv_range_nosync(unsigned long iova, size_t size, > if (!IS_ENABLED(CONFIG_64BIT) || smmu->version == ARM_SMMU_V1) { > iova &= ~12UL; > iova |= ARM_SMMU_CB_ASID(cfg); > - writel_relaxed(iova, reg); > + do { > + writel_relaxed(iova, reg); > + iova += granule; > + } while (size -= granule); > #ifdef CONFIG_64BIT > } else { > iova >>= 12; > iova |= (u64)ARM_SMMU_CB_ASID(cfg) << 48; > - writeq_relaxed(iova, reg); > + do { > + writeq_relaxed(iova, reg); > + iova += granule >> 12; > + } while (size -= granule) This doesn't compile. > #endif > } > #ifdef CONFIG_64BIT > @@ -610,7 +616,11 @@ static void arm_smmu_tlb_inv_range_nosync(unsigned long iova, size_t size, > reg = ARM_SMMU_CB_BASE(smmu) + ARM_SMMU_CB(smmu, cfg->cbndx); > reg += leaf ? ARM_SMMU_CB_S2_TLBIIPAS2L : > ARM_SMMU_CB_S2_TLBIIPAS2; > - writeq_relaxed(iova >> 12, reg); > + iova >>= 12; > + do { > + writeq_relaxed(iova, reg); > + iova += granule >> 12; > + } while (size -= granule) Same here. Please at least build your patches, and preferably test them too. Will
next prev parent reply other threads:[~2015-12-07 18:28 UTC|newest] Thread overview: 40+ messages / expand[flat|nested] mbox.gz Atom feed top 2015-12-04 17:52 [PATCH 0/5] io-pgtable fixes + ARM short-descriptor format Robin Murphy 2015-12-04 17:52 ` Robin Murphy [not found] ` <cover.1449246988.git.robin.murphy-5wv7dgnIgG8@public.gmane.org> 2015-12-04 17:52 ` [PATCH 1/5] iommu/io-pgtable-arm: Avoid dereferencing bogus PTEs Robin Murphy 2015-12-04 17:52 ` Robin Murphy [not found] ` <ad5898fd59575d0e2a8dccabafde71650f44e2a8.1449246988.git.robin.murphy-5wv7dgnIgG8@public.gmane.org> 2015-12-13 21:41 ` Laurent Pinchart 2015-12-13 21:41 ` Laurent Pinchart 2015-12-14 15:33 ` Robin Murphy 2015-12-14 15:33 ` Robin Murphy 2015-12-04 17:52 ` [PATCH 2/5] iommu/io-pgtable: Indicate granule for TLB maintenance Robin Murphy 2015-12-04 17:52 ` Robin Murphy [not found] ` <67223d4b1ff57f3f46e8c3102e663a063a50a7f7.1449246988.git.robin.murphy-5wv7dgnIgG8@public.gmane.org> 2015-12-07 11:08 ` Will Deacon 2015-12-07 11:08 ` Will Deacon [not found] ` <20151207110804.GA23430-5wv7dgnIgG8@public.gmane.org> 2015-12-07 12:09 ` Robin Murphy 2015-12-07 12:09 ` Robin Murphy [not found] ` <56657714.50504-5wv7dgnIgG8@public.gmane.org> 2015-12-07 13:48 ` Will Deacon 2015-12-07 13:48 ` Will Deacon 2015-12-07 18:18 ` [PATCH v2] " Robin Murphy 2015-12-07 18:18 ` Robin Murphy 2015-12-04 17:53 ` [PATCH 3/5] iommu/arm-smmu: Invalidate TLBs properly Robin Murphy 2015-12-04 17:53 ` Robin Murphy [not found] ` <2acaea8656f14a4421d7d466dd242fe5a3d0f6f6.1449246988.git.robin.murphy-5wv7dgnIgG8@public.gmane.org> 2015-12-07 11:09 ` Will Deacon 2015-12-07 11:09 ` Will Deacon [not found] ` <20151207110939.GB23430-5wv7dgnIgG8@public.gmane.org> 2015-12-07 13:09 ` Robin Murphy 2015-12-07 13:09 ` Robin Murphy [not found] ` <5665850F.1060406-5wv7dgnIgG8@public.gmane.org> 2015-12-07 13:34 ` Will Deacon 2015-12-07 13:34 ` Will Deacon 2015-12-07 18:18 ` [PATCH v2] " Robin Murphy 2015-12-07 18:18 ` Robin Murphy [not found] ` <ac2d6aedf473cf01eb1df48a1f81614f0f74b0b1.1449501523.git.robin.murphy-5wv7dgnIgG8@public.gmane.org> 2015-12-07 18:28 ` Will Deacon [this message] 2015-12-07 18:28 ` Will Deacon 2015-12-04 17:53 ` [PATCH 4/5] iommu/io-pgtable: Make io_pgtable_ops_to_pgtable() macro common Robin Murphy 2015-12-04 17:53 ` Robin Murphy [not found] ` <ef5954ba727840a020b62b0135a1ce9f4a10fb2c.1449246988.git.robin.murphy-5wv7dgnIgG8@public.gmane.org> 2015-12-13 21:52 ` Laurent Pinchart 2015-12-13 21:52 ` Laurent Pinchart 2015-12-04 17:53 ` [PATCH 5/5] iommu/io-pgtable: Add ARMv7 short descriptor support Robin Murphy 2015-12-04 17:53 ` Robin Murphy [not found] ` <3c72de1e8caa28cbfd423de41c6cba812db4e7db.1449246988.git.robin.murphy-5wv7dgnIgG8@public.gmane.org> 2015-12-08 8:58 ` Yong Wu 2015-12-08 8:58 ` Yong Wu 2015-12-17 20:12 ` Robin Murphy 2015-12-17 20:12 ` Robin Murphy
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=20151207182821.GH26191@arm.com \ --to=will.deacon-5wv7dgnigg8@public.gmane.org \ --cc=iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org \ --cc=linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org \ --cc=robin.murphy-5wv7dgnIgG8@public.gmane.org \ /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: linkBe 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.