From: Andrew Morton Subject: [patch 3/7] mm/gup: fix fixup_user_fault() on multiple retries Date: Wed, 13 May 2020 17:50:41 -0700 Message-ID: <20200514005041.Biir9x7Bl%akpm@linux-foundation.org> References: <20200513175005.1f4839360c18c0238df292d1@linux-foundation.org> Reply-To: linux-kernel@vger.kernel.org Return-path: Received: from mail.kernel.org ([198.145.29.99]:55922 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1728313AbgENAum (ORCPT ); Wed, 13 May 2020 20:50:42 -0400 In-Reply-To: <20200513175005.1f4839360c18c0238df292d1@linux-foundation.org> Sender: mm-commits-owner@vger.kernel.org List-Id: mm-commits@vger.kernel.org To: akpm@linux-foundation.org, alex.williamson@redhat.com, bgeffon@google.com, linux-mm@kvack.org, mm-commits@vger.kernel.org, peterx@redhat.com, torvalds@linux-foundation.org From: Peter Xu Subject: mm/gup: fix fixup_user_fault() on multiple retries This part is overlooked when reworking the gup code on multiple retries. When we get the 2nd+ retry, we'll be with TRIED flag set. Current code will bail out on the 2nd retry because the !TRIED check will fail so the retry logic will be skipped. What's worse is that, it will also return zero which errornously hints the caller that the page is faulted in while it's not. The !TRIED flag check seems to not be needed even before the mutliple retries change because if we get a VM_FAULT_RETRY, it must be the 1st retry, and we should not have TRIED set for that. Fix it by removing the !TRIED check, at the meantime check against fatal signals properly before the page fault so we can still properly respond to the user killing the process during retries. Link: http://lkml.kernel.org/r/20200502003523.8204-1-peterx@redhat.com Fixes: 4426e945df58 ("mm/gup: allow VM_FAULT_RETRY for multiple times") Signed-off-by: Peter Xu Cc: Alex Williamson Cc: Brian Geffon Signed-off-by: Andrew Morton --- mm/gup.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) --- a/mm/gup.c~mm-gup-fix-fixup_user_fault-on-multiple-retries +++ a/mm/gup.c @@ -1218,6 +1218,10 @@ retry: if (!vma_permits_fault(vma, fault_flags)) return -EFAULT; + if ((fault_flags & FAULT_FLAG_KILLABLE) && + fatal_signal_pending(current)) + return -EINTR; + ret = handle_mm_fault(vma, address, fault_flags); major |= ret & VM_FAULT_MAJOR; if (ret & VM_FAULT_ERROR) { @@ -1230,11 +1234,9 @@ retry: if (ret & VM_FAULT_RETRY) { down_read(&mm->mmap_sem); - if (!(fault_flags & FAULT_FLAG_TRIED)) { - *unlocked = true; - fault_flags |= FAULT_FLAG_TRIED; - goto retry; - } + *unlocked = true; + fault_flags |= FAULT_FLAG_TRIED; + goto retry; } if (tsk) { _