[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <20190724065258.16603-8-hch@lst.de>
Date: Wed, 24 Jul 2019 08:52:58 +0200
From: Christoph Hellwig <hch@....de>
To: Jérôme Glisse <jglisse@...hat.com>,
Jason Gunthorpe <jgg@...lanox.com>,
Ben Skeggs <bskeggs@...hat.com>
Cc: Ralph Campbell <rcampbell@...dia.com>, linux-mm@...ck.org,
nouveau@...ts.freedesktop.org, dri-devel@...ts.freedesktop.org,
linux-kernel@...r.kernel.org
Subject: [PATCH 7/7] mm: comment on VM_FAULT_RETRY semantics in handle_mm_fault
From: Jason Gunthorpe <jgg@...lanox.com>
The magic dropping of mmap_sem when handle_mm_fault returns
VM_FAULT_RETRY is rather subtile. Add a comment explaining it.
Signed-off-by: Jason Gunthorpe <jgg@...lanox.com>
[hch: wrote a changelog]
Signed-off-by: Christoph Hellwig <hch@....de>
---
mm/hmm.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/mm/hmm.c b/mm/hmm.c
index 16b6731a34db..54b3a4162ae9 100644
--- a/mm/hmm.c
+++ b/mm/hmm.c
@@ -301,8 +301,10 @@ static int hmm_vma_do_fault(struct mm_walk *walk, unsigned long addr,
flags |= hmm_vma_walk->block ? 0 : FAULT_FLAG_ALLOW_RETRY;
flags |= write_fault ? FAULT_FLAG_WRITE : 0;
ret = handle_mm_fault(vma, addr, flags);
- if (ret & VM_FAULT_RETRY)
+ if (ret & VM_FAULT_RETRY) {
+ /* Note, handle_mm_fault did up_read(&mm->mmap_sem)) */
return -EAGAIN;
+ }
if (ret & VM_FAULT_ERROR) {
*pfn = range->values[HMM_PFN_ERROR];
return -EFAULT;
--
2.20.1
Powered by blists - more mailing lists