[PATCHv3 0/5] fix xfs by making I/O to vmap/vmalloc areas work

[PATCHv3 0/5] fix xfs by making I/O to vmap/vmalloc areas work

[James Bottomley]

From: James Bottomley <James.Bottomley@HansenPartnership.com>

This is essentially a small tidy up from the previous series.  I
thought about the Ben H additions, but since power doesn't seem to
need this, they seemed a bit moot (we can expand the API when an
actual user comes along).

The patch series adds a flush/invalidate_kernel_vmap_range() API that
drivers using vmap/vmalloc areas must use before sending tohse areas
for I/O.  This makes it crystal clear that coherence on these areas is
the responsibility of the driver alone.  Fortunately xfs is the only
thing in the kernel actually doing I/O to vmap areas.

Sin ce xfs is completely broken on most VIPT architectures without
this, I'd like to submit it as a bug fix for 2.6.33.  Unfortunately,
we actually have some parisc xfs users whose data is curently at
severe risk.

James

---

James Bottomley (5):
  mm: add coherence API for DMA to vmalloc/vmap areas
  parisc: add mm API for DMA to vmalloc/vmap areas
  arm: add mm API for DMA to vmalloc/vmap areas
  sh: add mm API for DMA to vmalloc/vmap areas
  xfs: fix xfs to work with Virtually Indexed architectures

 Documentation/cachetlb.txt           |   24 ++++++++++++++++++++++++
 arch/arm/include/asm/cacheflush.h    |   10 ++++++++++
 arch/parisc/include/asm/cacheflush.h |   12 ++++++++++++
 arch/sh/include/asm/cacheflush.h     |    8 ++++++++
 fs/xfs/linux-2.6/xfs_buf.c           |   30 +++++++++++++++++++++++++++++-
 include/linux/highmem.h              |    6 ++++++
 6 files changed, 89 insertions(+), 1 deletions(-)

[PATCHv3 1/5] mm: add coherence API for DMA to vmalloc/vmap areas

[James Bottomley]

On Virtually Indexed architectures (which don't do automatic alias
resolution in their caches), we have to flush via the correct
virtual address to prepare pages for DMA.  On some architectures
(like arm) we cannot prevent the CPU from doing data movein along
the alias (and thus giving stale read data), so we not only have to
introduce a flush API to push dirty cache lines out, but also an invalidate
API to kill inconsistent cache lines that may have moved in before
DMA changed the data

Signed-off-by: James Bottomley <James.Bottomley@suse.de>
---
 Documentation/cachetlb.txt |   24 ++++++++++++++++++++++++
 include/linux/highmem.h    |    6 ++++++
 2 files changed, 30 insertions(+), 0 deletions(-)

diff --git a/Documentation/cachetlb.txt b/Documentation/cachetlb.txt
index da42ab4..b231414 100644
--- a/Documentation/cachetlb.txt
+++ b/Documentation/cachetlb.txt
@@ -377,3 +377,27 @@ maps this page at its virtual address.
 	All the functionality of flush_icache_page can be implemented in
 	flush_dcache_page and update_mmu_cache. In 2.7 the hope is to
 	remove this interface completely.
+
+The final category of APIs is for I/O to deliberately aliased address
+ranges inside the kernel.  Such aliases are set up by use of the
+vmap/vmalloc API.  Since kernel I/O goes via physical pages, the I/O
+subsystem assumes that the user mapping and kernel offset mapping are
+the only aliases.  This isn't true for vmap aliases, so anything in
+the kernel trying to do I/O to vmap areas must manually manage
+coherency.  It must do this by flushing the vmap range before doing
+I/O and invalidating it after the I/O returns.
+
+  void flush_kernel_vmap_range(void *vaddr, int size)
+       flushes the kernel cache for a given virtual address range in
+       the vmap area.  This is to make sure that any data the kernel
+       modified in the vmap range is made visible to the physical
+       page.  The design is to make this area safe to perform I/O on.
+       Note that this API does *not* also flush the offset map alias
+       of the area.
+
+  void invalidate_kernel_vmap_range(void *vaddr, int size) invalidates
+       the cache for a given virtual address range in the vmap area
+       which prevents the processor from making the cache stale by
+       speculatively reading data while the I/O was occurring to the
+       physical pages.  This is only necessary for data reads into the
+       vmap area.
diff --git a/include/linux/highmem.h b/include/linux/highmem.h
index 211ff44..adfe101 100644
--- a/include/linux/highmem.h
+++ b/include/linux/highmem.h
@@ -17,6 +17,12 @@ static inline void flush_anon_page(struct vm_area_struct *vma, struct page *page
 static inline void flush_kernel_dcache_page(struct page *page)
 {
 }
+static inline void flush_kernel_vmap_range(void *vaddr, int size)
+{
+}
+static inline void invalidate_kernel_vmap_range(void *vaddr, int size)
+{
+}
 #endif
 
 #include <asm/kmap_types.h>
-- 
1.6.5

[PATCHv3 2/5] parisc: add mm API for DMA to vmalloc/vmap areas

[James Bottomley]

We already have an API to flush a kernel page along an alias
address, so use it.  The TLB purge prevents the CPU from doing
speculative moveins on the flushed address, so we don't need to
implement and invalidate.

Signed-off-by: James Bottomley <James.Bottomley@suse.de>
---
 arch/parisc/include/asm/cacheflush.h |   12 ++++++++++++
 1 files changed, 12 insertions(+), 0 deletions(-)

diff --git a/arch/parisc/include/asm/cacheflush.h b/arch/parisc/include/asm/cacheflush.h
index 7a73b61..4772777 100644
--- a/arch/parisc/include/asm/cacheflush.h
+++ b/arch/parisc/include/asm/cacheflush.h
@@ -38,6 +38,18 @@ void flush_cache_mm(struct mm_struct *mm);
 
 #define flush_kernel_dcache_range(start,size) \
 	flush_kernel_dcache_range_asm((start), (start)+(size));
+/* vmap range flushes and invalidates.  Architecturally, we don't need
+ * the invalidate, because the CPU should refuse to speculate once an
+ * area has been flushed, so invalidate is left empty */
+static inline void flush_kernel_vmap_range(void *vaddr, int size)
+{
+	unsigned long start = (unsigned long)vaddr;
+
+	flush_kernel_dcache_range_asm(start, start + size);
+}
+static inline void invalidate_kernel_vmap_range(void *vaddr, int size)
+{
+}
 
 #define flush_cache_vmap(start, end)		flush_cache_all()
 #define flush_cache_vunmap(start, end)		flush_cache_all()
-- 
1.6.5

[PATCHv3 2/5] parisc: add mm API for DMA to vmalloc/vmap areas

[James Bottomley]

We already have an API to flush a kernel page along an alias
address, so use it.  The TLB purge prevents the CPU from doing
speculative moveins on the flushed address, so we don't need to
implement and invalidate.

Signed-off-by: James Bottomley <James.Bottomley@suse.de>
---
 arch/parisc/include/asm/cacheflush.h |   12 ++++++++++++
 1 files changed, 12 insertions(+), 0 deletions(-)

diff --git a/arch/parisc/include/asm/cacheflush.h b/arch/parisc/include/asm/cacheflush.h
index 7a73b61..4772777 100644
--- a/arch/parisc/include/asm/cacheflush.h
+++ b/arch/parisc/include/asm/cacheflush.h
@@ -38,6 +38,18 @@ void flush_cache_mm(struct mm_struct *mm);
 
 #define flush_kernel_dcache_range(start,size) \
 	flush_kernel_dcache_range_asm((start), (start)+(size));
+/* vmap range flushes and invalidates.  Architecturally, we don't need
+ * the invalidate, because the CPU should refuse to speculate once an
+ * area has been flushed, so invalidate is left empty */
+static inline void flush_kernel_vmap_range(void *vaddr, int size)
+{
+	unsigned long start = (unsigned long)vaddr;
+
+	flush_kernel_dcache_range_asm(start, start + size);
+}
+static inline void invalidate_kernel_vmap_range(void *vaddr, int size)
+{
+}
 
 #define flush_cache_vmap(start, end)		flush_cache_all()
 #define flush_cache_vunmap(start, end)		flush_cache_all()
-- 
1.6.5

[PATCHv3 3/5] arm: add mm API for DMA to vmalloc/vmap areas

[James Bottomley]

ARM cannot prevent cache movein, so this patch implements both the
flush and invalidate pieces of the API.

Signed-off-by: James Bottomley <James.Bottomley@suse.de>
---
 arch/arm/include/asm/cacheflush.h |   10 ++++++++++
 1 files changed, 10 insertions(+), 0 deletions(-)

diff --git a/arch/arm/include/asm/cacheflush.h b/arch/arm/include/asm/cacheflush.h
index 730aefc..4ae503c 100644
--- a/arch/arm/include/asm/cacheflush.h
+++ b/arch/arm/include/asm/cacheflush.h
@@ -432,6 +432,16 @@ static inline void __flush_icache_all(void)
 	    : "r" (0));
 #endif
 }
+static inline void flush_kernel_vmap_range(void *addr, int size)
+{
+	if ((cache_is_vivt() || cache_is_vipt_aliasing()))
+	  __cpuc_flush_dcache_area(addr, (size_t)size);
+}
+static inline void invalidate_kernel_vmap_range(void *addr, int size)
+{
+	if ((cache_is_vivt() || cache_is_vipt_aliasing()))
+	  __cpuc_flush_dcache_area(addr, (size_t)size);
+}
 
 #define ARCH_HAS_FLUSH_ANON_PAGE
 static inline void flush_anon_page(struct vm_area_struct *vma,
-- 
1.6.5

[PATCHv3 3/5] arm: add mm API for DMA to vmalloc/vmap areas

[James Bottomley]

ARM cannot prevent cache movein, so this patch implements both the
flush and invalidate pieces of the API.

Signed-off-by: James Bottomley <James.Bottomley@suse.de>
---
 arch/arm/include/asm/cacheflush.h |   10 ++++++++++
 1 files changed, 10 insertions(+), 0 deletions(-)

diff --git a/arch/arm/include/asm/cacheflush.h b/arch/arm/include/asm/cacheflush.h
index 730aefc..4ae503c 100644
--- a/arch/arm/include/asm/cacheflush.h
+++ b/arch/arm/include/asm/cacheflush.h
@@ -432,6 +432,16 @@ static inline void __flush_icache_all(void)
 	    : "r" (0));
 #endif
 }
+static inline void flush_kernel_vmap_range(void *addr, int size)
+{
+	if ((cache_is_vivt() || cache_is_vipt_aliasing()))
+	  __cpuc_flush_dcache_area(addr, (size_t)size);
+}
+static inline void invalidate_kernel_vmap_range(void *addr, int size)
+{
+	if ((cache_is_vivt() || cache_is_vipt_aliasing()))
+	  __cpuc_flush_dcache_area(addr, (size_t)size);
+}
 
 #define ARCH_HAS_FLUSH_ANON_PAGE
 static inline void flush_anon_page(struct vm_area_struct *vma,
-- 
1.6.5

[PATCHv3 4/5] sh: add mm API for DMA to vmalloc/vmap areas

[James Bottomley]

Signed-off-by: James Bottomley <James.Bottomley@suse.de>
---
 arch/sh/include/asm/cacheflush.h |    8 ++++++++
 1 files changed, 8 insertions(+), 0 deletions(-)

diff --git a/arch/sh/include/asm/cacheflush.h b/arch/sh/include/asm/cacheflush.h
index dda96eb..da3ebec 100644
--- a/arch/sh/include/asm/cacheflush.h
+++ b/arch/sh/include/asm/cacheflush.h
@@ -63,6 +63,14 @@ static inline void flush_anon_page(struct vm_area_struct *vma,
 	if (boot_cpu_data.dcache.n_aliases && PageAnon(page))
 		__flush_anon_page(page, vmaddr);
 }
+static inline void flush_kernel_vmap_range(void *addr, int size)
+{
+	__flush_wback_region(addr, size);
+}
+static inline void invalidate_kernel_vmap_range(void *addr, int size)
+{
+	__flush_invalidate_region(addr, size);
+}
 
 #define ARCH_HAS_FLUSH_KERNEL_DCACHE_PAGE
 static inline void flush_kernel_dcache_page(struct page *page)
-- 
1.6.5

[PATCHv3 4/5] sh: add mm API for DMA to vmalloc/vmap areas

[James Bottomley]

Signed-off-by: James Bottomley <James.Bottomley@suse.de>
---
 arch/sh/include/asm/cacheflush.h |    8 ++++++++
 1 files changed, 8 insertions(+), 0 deletions(-)

diff --git a/arch/sh/include/asm/cacheflush.h b/arch/sh/include/asm/cacheflush.h
index dda96eb..da3ebec 100644
--- a/arch/sh/include/asm/cacheflush.h
+++ b/arch/sh/include/asm/cacheflush.h
@@ -63,6 +63,14 @@ static inline void flush_anon_page(struct vm_area_struct *vma,
 	if (boot_cpu_data.dcache.n_aliases && PageAnon(page))
 		__flush_anon_page(page, vmaddr);
 }
+static inline void flush_kernel_vmap_range(void *addr, int size)
+{
+	__flush_wback_region(addr, size);
+}
+static inline void invalidate_kernel_vmap_range(void *addr, int size)
+{
+	__flush_invalidate_region(addr, size);
+}
 
 #define ARCH_HAS_FLUSH_KERNEL_DCACHE_PAGE
 static inline void flush_kernel_dcache_page(struct page *page)
-- 
1.6.5

[PATCHv3 5/5] xfs: fix xfs to work with Virtually Indexed architectures

[James Bottomley]

xfs_buf.c includes what is essentially a hand rolled version of
blk_rq_map_kern().  In order to work properly with the vmalloc buffers
that xfs uses, this hand rolled routine must also implement the flushing
API for vmap/vmalloc areas.

[style updates from hch@lst.de]
Signed-off-by: James Bottomley <James.Bottomley@suse.de>
---
 fs/xfs/linux-2.6/xfs_buf.c |   30 +++++++++++++++++++++++++++++-
 1 files changed, 29 insertions(+), 1 deletions(-)

diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c
index 77b8be8..6f3ebb6 100644
--- a/fs/xfs/linux-2.6/xfs_buf.c
+++ b/fs/xfs/linux-2.6/xfs_buf.c
@@ -76,6 +76,27 @@ struct workqueue_struct *xfsconvertd_workqueue;
 #define xfs_buf_deallocate(bp) \
 	kmem_zone_free(xfs_buf_zone, (bp));
 
+static inline int
+xfs_buf_is_vmapped(
+	struct xfs_buf	*bp)
+{
+	/*
+	 * Return true if the buffer is vmapped.
+	 *
+	 * The XBF_MAPPED flag is set if the buffer should be mapped, but the
+	 * code is clever enough to know it doesn't have to map a single page,
+	 * so the check has to be both for XBF_MAPPED and bp->b_page_count > 1.
+	 */
+	return (bp->b_flags & XBF_MAPPED) && bp->b_page_count > 1;
+}
+
+static inline int
+xfs_buf_vmap_len(
+	struct xfs_buf	*bp)
+{
+	return (bp->b_page_count * PAGE_SIZE) - bp->b_offset;
+}
+
 /*
  *	Page Region interfaces.
  *
@@ -314,7 +335,7 @@ xfs_buf_free(
 	if (bp->b_flags & (_XBF_PAGE_CACHE|_XBF_PAGES)) {
 		uint		i;
 
-		if ((bp->b_flags & XBF_MAPPED) && (bp->b_page_count > 1))
+		if (xfs_buf_is_vmapped(bp))
 			free_address(bp->b_addr - bp->b_offset);
 
 		for (i = 0; i < bp->b_page_count; i++) {
@@ -1107,6 +1128,9 @@ xfs_buf_bio_end_io(
 
 	xfs_buf_ioerror(bp, -error);
 
+	if (!error && xfs_buf_is_vmapped(bp) && (bp->b_flags & XBF_READ))
+		invalidate_kernel_vmap_range(bp->b_addr, xfs_buf_vmap_len(bp));
+
 	do {
 		struct page	*page = bvec->bv_page;
 
@@ -1216,6 +1240,10 @@ next_chunk:
 
 submit_io:
 	if (likely(bio->bi_size)) {
+		if (xfs_buf_is_vmapped(bp)) {
+			flush_kernel_vmap_range(bp->b_addr,
+						xfs_buf_vmap_len(bp));
+		}
 		submit_bio(rw, bio);
 		if (size)
 			goto next_chunk;
-- 
1.6.5

Re: [PATCHv3 0/5] fix xfs by making I/O to vmap/vmalloc areas work

[Christoph Hellwig]

On Fri, Feb 05, 2010 at 09:50:52AM -0600, James Bottomley wrote:
> From: James Bottomley <James.Bottomley@HansenPartnership.com>
> 
> This is essentially a small tidy up from the previous series.  I
> thought about the Ben H additions, but since power doesn't seem to
> need this, they seemed a bit moot (we can expand the API when an
> actual user comes along).
> 
> The patch series adds a flush/invalidate_kernel_vmap_range() API that
> drivers using vmap/vmalloc areas must use before sending tohse areas
> for I/O.  This makes it crystal clear that coherence on these areas is
> the responsibility of the driver alone.  Fortunately xfs is the only
> thing in the kernel actually doing I/O to vmap areas.
> 
> Sin ce xfs is completely broken on most VIPT architectures without
> this, I'd like to submit it as a bug fix for 2.6.33.  Unfortunately,
> we actually have some parisc xfs users whose data is curently at
> severe risk.

Agreed.  There's also a lot of ARM users popping up with this recently,
while others worked around it using local flushing hacks previously.

Re: [PATCHv3 0/5] fix xfs by making I/O to vmap/vmalloc areas work

[Kyle McMartin]

On Fri, Feb 05, 2010 at 05:06:40PM +0100, Christoph Hellwig wrote:
> > From: James Bottomley <James.Bottomley@HansenPartnership.com>
> > 
> 
> Agreed.  There's also a lot of ARM users popping up with this recently,
> while others worked around it using local flushing hacks previously.
> 

Thanks, I've applied these to parisc/next to give them a shakedown in
linux-next.

regards, Kyle.

Re: [PATCHv3 0/5] fix xfs by making I/O to vmap/vmalloc areas work

[Kyle McMartin]

On Fri, Feb 05, 2010 at 05:06:40PM +0100, Christoph Hellwig wrote:
> > From: James Bottomley <James.Bottomley@HansenPartnership.com>
> > 
> 
> Agreed.  There's also a lot of ARM users popping up with this recently,
> while others worked around it using local flushing hacks previously.
> 

Thanks, I've applied these to parisc/next to give them a shakedown in
linux-next.

regards, Kyle.