<feed xmlns='http://www.w3.org/2005/Atom'>
<title>linux-toradex.git/kernel/dma/contiguous.c, branch v5.10-rc1</title>
<subtitle>Linux kernel for Apalis and Colibri modules</subtitle>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/'/>
<entry>
<title>Merge tag 'dma-mapping-5.10' of git://git.infradead.org/users/hch/dma-mapping</title>
<updated>2020-10-15T21:43:29+00:00</updated>
<author>
<name>Linus Torvalds</name>
<email>torvalds@linux-foundation.org</email>
</author>
<published>2020-10-15T21:43:29+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=5a32c3413d3340f90c82c84b375ad4b335a59f28'/>
<id>5a32c3413d3340f90c82c84b375ad4b335a59f28</id>
<content type='text'>
Pull dma-mapping updates from Christoph Hellwig:

 - rework the non-coherent DMA allocator

 - move private definitions out of &lt;linux/dma-mapping.h&gt;

 - lower CMA_ALIGNMENT (Paul Cercueil)

 - remove the omap1 dma address translation in favor of the common code

 - make dma-direct aware of multiple dma offset ranges (Jim Quinlan)

 - support per-node DMA CMA areas (Barry Song)

 - increase the default seg boundary limit (Nicolin Chen)

 - misc fixes (Robin Murphy, Thomas Tai, Xu Wang)

 - various cleanups

* tag 'dma-mapping-5.10' of git://git.infradead.org/users/hch/dma-mapping: (63 commits)
  ARM/ixp4xx: add a missing include of dma-map-ops.h
  dma-direct: simplify the DMA_ATTR_NO_KERNEL_MAPPING handling
  dma-direct: factor out a dma_direct_alloc_from_pool helper
  dma-direct check for highmem pages in dma_direct_alloc_pages
  dma-mapping: merge &lt;linux/dma-noncoherent.h&gt; into &lt;linux/dma-map-ops.h&gt;
  dma-mapping: move large parts of &lt;linux/dma-direct.h&gt; to kernel/dma
  dma-mapping: move dma-debug.h to kernel/dma/
  dma-mapping: remove &lt;asm/dma-contiguous.h&gt;
  dma-mapping: merge &lt;linux/dma-contiguous.h&gt; into &lt;linux/dma-map-ops.h&gt;
  dma-contiguous: remove dma_contiguous_set_default
  dma-contiguous: remove dev_set_cma_area
  dma-contiguous: remove dma_declare_contiguous
  dma-mapping: split &lt;linux/dma-mapping.h&gt;
  cma: decrease CMA_ALIGNMENT lower limit to 2
  firewire-ohci: use dma_alloc_pages
  dma-iommu: implement -&gt;alloc_noncoherent
  dma-mapping: add new {alloc,free}_noncoherent dma_map_ops methods
  dma-mapping: add a new dma_alloc_pages API
  dma-mapping: remove dma_cache_sync
  53c700: convert to dma_alloc_noncoherent
  ...
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Pull dma-mapping updates from Christoph Hellwig:

 - rework the non-coherent DMA allocator

 - move private definitions out of &lt;linux/dma-mapping.h&gt;

 - lower CMA_ALIGNMENT (Paul Cercueil)

 - remove the omap1 dma address translation in favor of the common code

 - make dma-direct aware of multiple dma offset ranges (Jim Quinlan)

 - support per-node DMA CMA areas (Barry Song)

 - increase the default seg boundary limit (Nicolin Chen)

 - misc fixes (Robin Murphy, Thomas Tai, Xu Wang)

 - various cleanups

* tag 'dma-mapping-5.10' of git://git.infradead.org/users/hch/dma-mapping: (63 commits)
  ARM/ixp4xx: add a missing include of dma-map-ops.h
  dma-direct: simplify the DMA_ATTR_NO_KERNEL_MAPPING handling
  dma-direct: factor out a dma_direct_alloc_from_pool helper
  dma-direct check for highmem pages in dma_direct_alloc_pages
  dma-mapping: merge &lt;linux/dma-noncoherent.h&gt; into &lt;linux/dma-map-ops.h&gt;
  dma-mapping: move large parts of &lt;linux/dma-direct.h&gt; to kernel/dma
  dma-mapping: move dma-debug.h to kernel/dma/
  dma-mapping: remove &lt;asm/dma-contiguous.h&gt;
  dma-mapping: merge &lt;linux/dma-contiguous.h&gt; into &lt;linux/dma-map-ops.h&gt;
  dma-contiguous: remove dma_contiguous_set_default
  dma-contiguous: remove dev_set_cma_area
  dma-contiguous: remove dma_declare_contiguous
  dma-mapping: split &lt;linux/dma-mapping.h&gt;
  cma: decrease CMA_ALIGNMENT lower limit to 2
  firewire-ohci: use dma_alloc_pages
  dma-iommu: implement -&gt;alloc_noncoherent
  dma-mapping: add new {alloc,free}_noncoherent dma_map_ops methods
  dma-mapping: add a new dma_alloc_pages API
  dma-mapping: remove dma_cache_sync
  53c700: convert to dma_alloc_noncoherent
  ...
</pre>
</div>
</content>
</entry>
<entry>
<title>dma-contiguous: simplify cma_early_percent_memory()</title>
<updated>2020-10-14T01:38:35+00:00</updated>
<author>
<name>Mike Rapoport</name>
<email>rppt@linux.ibm.com</email>
</author>
<published>2020-10-13T23:57:22+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=e9aa36ccbb4e845d043bb51de274ae877ae925a7'/>
<id>e9aa36ccbb4e845d043bb51de274ae877ae925a7</id>
<content type='text'>
The memory size calculation in cma_early_percent_memory() traverses
memblock.memory rather than simply call memblock_phys_mem_size().  The
comment in that function suggests that at some point there should have
been call to memblock_analyze() before memblock_phys_mem_size() could be
used.  As of now, there is no memblock_analyze() at all and
memblock_phys_mem_size() can be used as soon as cold-plug memory is
registered with memblock.

Replace loop over memblock.memory with a call to memblock_phys_mem_size().

Signed-off-by: Mike Rapoport &lt;rppt@linux.ibm.com&gt;
Signed-off-by: Andrew Morton &lt;akpm@linux-foundation.org&gt;
Reviewed-by: Christoph Hellwig &lt;hch@lst.de&gt;
Reviewed-by: Baoquan He &lt;bhe@redhat.com&gt;
Cc: Andy Lutomirski &lt;luto@kernel.org&gt;
Cc: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
Cc: Borislav Petkov &lt;bp@alien8.de&gt;
Cc: Catalin Marinas &lt;catalin.marinas@arm.com&gt;
Cc: Daniel Axtens &lt;dja@axtens.net&gt;
Cc: Dave Hansen &lt;dave.hansen@linux.intel.com&gt;
Cc: Emil Renner Berthing &lt;kernel@esmil.dk&gt;
Cc: Hari Bathini &lt;hbathini@linux.ibm.com&gt;
Cc: Ingo Molnar &lt;mingo@kernel.org&gt;
Cc: Ingo Molnar &lt;mingo@redhat.com&gt;
Cc: Jonathan Cameron &lt;Jonathan.Cameron@huawei.com&gt;
Cc: Marek Szyprowski &lt;m.szyprowski@samsung.com&gt;
Cc: Max Filippov &lt;jcmvbkbc@gmail.com&gt;
Cc: Michael Ellerman &lt;mpe@ellerman.id.au&gt;
Cc: Michal Simek &lt;monstr@monstr.eu&gt;
Cc: Miguel Ojeda &lt;miguel.ojeda.sandonis@gmail.com&gt;
Cc: Palmer Dabbelt &lt;palmer@dabbelt.com&gt;
Cc: Paul Mackerras &lt;paulus@samba.org&gt;
Cc: Paul Walmsley &lt;paul.walmsley@sifive.com&gt;
Cc: Peter Zijlstra &lt;peterz@infradead.org&gt;
Cc: Russell King &lt;linux@armlinux.org.uk&gt;
Cc: Stafford Horne &lt;shorne@gmail.com&gt;
Cc: Thomas Bogendoerfer &lt;tsbogend@alpha.franken.de&gt;
Cc: Thomas Gleixner &lt;tglx@linutronix.de&gt;
Cc: Will Deacon &lt;will@kernel.org&gt;
Cc: Yoshinori Sato &lt;ysato@users.sourceforge.jp&gt;
Link: https://lkml.kernel.org/r/20200818151634.14343-3-rppt@kernel.org
Signed-off-by: Linus Torvalds &lt;torvalds@linux-foundation.org&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
The memory size calculation in cma_early_percent_memory() traverses
memblock.memory rather than simply call memblock_phys_mem_size().  The
comment in that function suggests that at some point there should have
been call to memblock_analyze() before memblock_phys_mem_size() could be
used.  As of now, there is no memblock_analyze() at all and
memblock_phys_mem_size() can be used as soon as cold-plug memory is
registered with memblock.

Replace loop over memblock.memory with a call to memblock_phys_mem_size().

Signed-off-by: Mike Rapoport &lt;rppt@linux.ibm.com&gt;
Signed-off-by: Andrew Morton &lt;akpm@linux-foundation.org&gt;
Reviewed-by: Christoph Hellwig &lt;hch@lst.de&gt;
Reviewed-by: Baoquan He &lt;bhe@redhat.com&gt;
Cc: Andy Lutomirski &lt;luto@kernel.org&gt;
Cc: Benjamin Herrenschmidt &lt;benh@kernel.crashing.org&gt;
Cc: Borislav Petkov &lt;bp@alien8.de&gt;
Cc: Catalin Marinas &lt;catalin.marinas@arm.com&gt;
Cc: Daniel Axtens &lt;dja@axtens.net&gt;
Cc: Dave Hansen &lt;dave.hansen@linux.intel.com&gt;
Cc: Emil Renner Berthing &lt;kernel@esmil.dk&gt;
Cc: Hari Bathini &lt;hbathini@linux.ibm.com&gt;
Cc: Ingo Molnar &lt;mingo@kernel.org&gt;
Cc: Ingo Molnar &lt;mingo@redhat.com&gt;
Cc: Jonathan Cameron &lt;Jonathan.Cameron@huawei.com&gt;
Cc: Marek Szyprowski &lt;m.szyprowski@samsung.com&gt;
Cc: Max Filippov &lt;jcmvbkbc@gmail.com&gt;
Cc: Michael Ellerman &lt;mpe@ellerman.id.au&gt;
Cc: Michal Simek &lt;monstr@monstr.eu&gt;
Cc: Miguel Ojeda &lt;miguel.ojeda.sandonis@gmail.com&gt;
Cc: Palmer Dabbelt &lt;palmer@dabbelt.com&gt;
Cc: Paul Mackerras &lt;paulus@samba.org&gt;
Cc: Paul Walmsley &lt;paul.walmsley@sifive.com&gt;
Cc: Peter Zijlstra &lt;peterz@infradead.org&gt;
Cc: Russell King &lt;linux@armlinux.org.uk&gt;
Cc: Stafford Horne &lt;shorne@gmail.com&gt;
Cc: Thomas Bogendoerfer &lt;tsbogend@alpha.franken.de&gt;
Cc: Thomas Gleixner &lt;tglx@linutronix.de&gt;
Cc: Will Deacon &lt;will@kernel.org&gt;
Cc: Yoshinori Sato &lt;ysato@users.sourceforge.jp&gt;
Link: https://lkml.kernel.org/r/20200818151634.14343-3-rppt@kernel.org
Signed-off-by: Linus Torvalds &lt;torvalds@linux-foundation.org&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>dma-mapping: remove &lt;asm/dma-contiguous.h&gt;</title>
<updated>2020-10-06T05:07:05+00:00</updated>
<author>
<name>Christoph Hellwig</name>
<email>hch@lst.de</email>
</author>
<published>2020-09-11T09:04:43+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=5db5d93089880c3cc9e83ca8bba68a5502e92dfe'/>
<id>5db5d93089880c3cc9e83ca8bba68a5502e92dfe</id>
<content type='text'>
Just provide a weak default definition of dma_contiguous_early_fixup and
let arm override it.

Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Just provide a weak default definition of dma_contiguous_early_fixup and
let arm override it.

Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>dma-mapping: merge &lt;linux/dma-contiguous.h&gt; into &lt;linux/dma-map-ops.h&gt;</title>
<updated>2020-10-06T05:07:04+00:00</updated>
<author>
<name>Christoph Hellwig</name>
<email>hch@lst.de</email>
</author>
<published>2020-09-11T08:56:52+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=0b1abd1fb7efafc25231c54a67c6fbb3d3127efd'/>
<id>0b1abd1fb7efafc25231c54a67c6fbb3d3127efd</id>
<content type='text'>
Merge dma-contiguous.h into dma-map-ops.h, after removing the comment
describing the contiguous allocator into kernel/dma/contigous.c.

Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Merge dma-contiguous.h into dma-map-ops.h, after removing the comment
describing the contiguous allocator into kernel/dma/contigous.c.

Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>dma-contiguous: remove dma_contiguous_set_default</title>
<updated>2020-10-06T05:07:03+00:00</updated>
<author>
<name>Christoph Hellwig</name>
<email>hch@lst.de</email>
</author>
<published>2020-09-11T08:56:40+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=580a0cc9c3f662e0b10136bc8af1e672e472806f'/>
<id>580a0cc9c3f662e0b10136bc8af1e672e472806f</id>
<content type='text'>
dma_contiguous_set_default contains a trivial assignment, and has a
single caller that is compiled if CONFIG_CMA_DMA is enabled.

Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
dma_contiguous_set_default contains a trivial assignment, and has a
single caller that is compiled if CONFIG_CMA_DMA is enabled.

Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>dma-contiguous: remove dev_set_cma_area</title>
<updated>2020-10-06T05:07:03+00:00</updated>
<author>
<name>Christoph Hellwig</name>
<email>hch@lst.de</email>
</author>
<published>2020-09-11T08:56:03+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=5af638931eb374aa0894d8343cee72f50307ef20'/>
<id>5af638931eb374aa0894d8343cee72f50307ef20</id>
<content type='text'>
dev_set_cma_area contains a trivial assignment.  It has just three
callers that all have a non-NULL device and depend on CONFIG_DMA_CMA,
so remove the wrapper.

Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
dev_set_cma_area contains a trivial assignment.  It has just three
callers that all have a non-NULL device and depend on CONFIG_DMA_CMA,
so remove the wrapper.

Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>mm: cma: use CMA_MAX_NAME to define the length of cma name array</title>
<updated>2020-09-01T07:19:43+00:00</updated>
<author>
<name>Barry Song</name>
<email>song.bao.hua@hisilicon.com</email>
</author>
<published>2020-08-23T23:03:09+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=2281f797f5524abb8fff66bf8540b4f4687332a2'/>
<id>2281f797f5524abb8fff66bf8540b4f4687332a2</id>
<content type='text'>
CMA_MAX_NAME should be visible to CMA's users as they might need it to set
the name of CMA areas and avoid hardcoding the size locally.
So this patch moves CMA_MAX_NAME from local header file to include/linux
header file and removes the hardcode in both hugetlb.c and contiguous.c.

Signed-off-by: Barry Song &lt;song.bao.hua@hisilicon.com&gt;
Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
CMA_MAX_NAME should be visible to CMA's users as they might need it to set
the name of CMA areas and avoid hardcoding the size locally.
So this patch moves CMA_MAX_NAME from local header file to include/linux
header file and removes the hardcode in both hugetlb.c and contiguous.c.

Signed-off-by: Barry Song &lt;song.bao.hua@hisilicon.com&gt;
Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>dma-contiguous: provide the ability to reserve per-numa CMA</title>
<updated>2020-09-01T07:19:28+00:00</updated>
<author>
<name>Barry Song</name>
<email>song.bao.hua@hisilicon.com</email>
</author>
<published>2020-08-23T23:03:07+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=b7176c261cdbced87bed9562577333150ed05b01'/>
<id>b7176c261cdbced87bed9562577333150ed05b01</id>
<content type='text'>
Right now, drivers like ARM SMMU are using dma_alloc_coherent() to get
coherent DMA buffers to save their command queues and page tables. As
there is only one default CMA in the whole system, SMMUs on nodes other
than node0 will get remote memory. This leads to significant latency.

This patch provides per-numa CMA so that drivers like SMMU can get local
memory. Tests show localizing CMA can decrease dma_unmap latency much.
For instance, before this patch, SMMU on node2  has to wait for more than
560ns for the completion of CMD_SYNC in an empty command queue; with this
patch, it needs 240ns only.

A positive side effect of this patch would be improving performance even
further for those users who are worried about performance more than DMA
security and use iommu.passthrough=1 to skip IOMMU. With local CMA, all
drivers can get local coherent DMA buffers.

Also, this patch changes the default CONFIG_CMA_AREAS to 19 in NUMA. As
1+CONFIG_CMA_AREAS should be quite enough for most servers on the market
even they enable both hugetlb_cma and pernuma_cma.
2 numa nodes: 2(hugetlb) + 2(pernuma) + 1(default global cma) = 5
4 numa nodes: 4(hugetlb) + 4(pernuma) + 1(default global cma) = 9
8 numa nodes: 8(hugetlb) + 8(pernuma) + 1(default global cma) = 17

Signed-off-by: Barry Song &lt;song.bao.hua@hisilicon.com&gt;
Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Right now, drivers like ARM SMMU are using dma_alloc_coherent() to get
coherent DMA buffers to save their command queues and page tables. As
there is only one default CMA in the whole system, SMMUs on nodes other
than node0 will get remote memory. This leads to significant latency.

This patch provides per-numa CMA so that drivers like SMMU can get local
memory. Tests show localizing CMA can decrease dma_unmap latency much.
For instance, before this patch, SMMU on node2  has to wait for more than
560ns for the completion of CMD_SYNC in an empty command queue; with this
patch, it needs 240ns only.

A positive side effect of this patch would be improving performance even
further for those users who are worried about performance more than DMA
security and use iommu.passthrough=1 to skip IOMMU. With local CMA, all
drivers can get local coherent DMA buffers.

Also, this patch changes the default CONFIG_CMA_AREAS to 19 in NUMA. As
1+CONFIG_CMA_AREAS should be quite enough for most servers on the market
even they enable both hugetlb_cma and pernuma_cma.
2 numa nodes: 2(hugetlb) + 2(pernuma) + 1(default global cma) = 5
4 numa nodes: 4(hugetlb) + 4(pernuma) + 1(default global cma) = 9
8 numa nodes: 8(hugetlb) + 8(pernuma) + 1(default global cma) = 17

Signed-off-by: Barry Song &lt;song.bao.hua@hisilicon.com&gt;
Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>dma-contiguous: cleanup dma_alloc_contiguous</title>
<updated>2020-07-28T11:42:15+00:00</updated>
<author>
<name>Christoph Hellwig</name>
<email>hch@lst.de</email>
</author>
<published>2020-07-22T14:33:43+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=274b3f7bf34415eed106e479e4815e897ce5d763'/>
<id>274b3f7bf34415eed106e479e4815e897ce5d763</id>
<content type='text'>
Split out a cma_alloc_aligned helper to deal with the "interesting"
calling conventions for cma_alloc, which then allows to the main
function to be written straight forward.  This also takes advantage
of the fact that NULL dev arguments have been gone from the DMA API
for a while.

Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
Reviewed-by: Nicolin Chen &lt;nicoleotsuka@gmail.com&gt;
Reviewed-by: Barry Song &lt;song.bao.hua@hisilicon.com&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Split out a cma_alloc_aligned helper to deal with the "interesting"
calling conventions for cma_alloc, which then allows to the main
function to be written straight forward.  This also takes advantage
of the fact that NULL dev arguments have been gone from the DMA API
for a while.

Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
Reviewed-by: Nicolin Chen &lt;nicoleotsuka@gmail.com&gt;
Reviewed-by: Barry Song &lt;song.bao.hua@hisilicon.com&gt;
</pre>
</div>
</content>
</entry>
<entry>
<title>dma-contiguous: fix comment for dma_release_from_contiguous</title>
<updated>2020-04-25T11:17:06+00:00</updated>
<author>
<name>Peter Collingbourne</name>
<email>pcc@google.com</email>
</author>
<published>2020-04-23T16:31:31+00:00</published>
<link rel='alternate' type='text/html' href='https://git.toradex.cn/cgit/linux-toradex.git/commit/?id=298f3db6ee690259927b105d5ad1079563361323'/>
<id>298f3db6ee690259927b105d5ad1079563361323</id>
<content type='text'>
Commit 90ae409f9eb3 ("dma-direct: fix zone selection
after an unaddressable CMA allocation") changed the logic in
dma_release_from_contiguous to remove the normal pages fallback path,
but did not update the comment. Fix that.

Signed-off-by: Peter Collingbourne &lt;pcc@google.com&gt;
Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</content>
<content type='xhtml'>
<div xmlns='http://www.w3.org/1999/xhtml'>
<pre>
Commit 90ae409f9eb3 ("dma-direct: fix zone selection
after an unaddressable CMA allocation") changed the logic in
dma_release_from_contiguous to remove the normal pages fallback path,
but did not update the comment. Fix that.

Signed-off-by: Peter Collingbourne &lt;pcc@google.com&gt;
Signed-off-by: Christoph Hellwig &lt;hch@lst.de&gt;
</pre>
</div>
</content>
</entry>
</feed>
