On Wed, Jun 21, 2017 at 04:23:26PM +0800, Zorro Lang wrote: > We met a kernel assertion failure recently as below: > > XFS: Assertion failed: tp->t_blk_res_used <= tp->t_blk_res, file: fs/xfs/xfs_trans.c, line: 309 > > The problem comes when the several IO vectors are copied in, and > it runs into page faults. Can you please provide more information in commit log? It's not so clear where the problem lies in right now. And there's a typo in subject line :) generic: test writev with pega fault when it processes iov ^^^^ page > > Signed-off-by: Zorro Lang <zlang@xxxxxxxxxx> > --- > > Hi, > > Although I'm trying to add writev/readv test to fsstress, and it has tiny > chance to cover this bug (very hard to reproduce). But I still think using > a special case to cover this part is better. > > Thanks, > Zorro > > .gitignore | 1 + > src/Makefile | 2 +- > src/writev_on_pagefault.c | 74 +++++++++++++++++++++++++++++++++++++++++++++++ > tests/generic/443 | 61 ++++++++++++++++++++++++++++++++++++++ > tests/generic/443.out | 2 ++ > tests/generic/group | 1 + > 6 files changed, 140 insertions(+), 1 deletion(-) > create mode 100644 src/writev_on_pagefault.c > create mode 100755 tests/generic/443 > create mode 100644 tests/generic/443.out > > diff --git a/.gitignore b/.gitignore > index 39664b0..043e87f 100644 > --- a/.gitignore > +++ b/.gitignore > @@ -130,6 +130,7 @@ > /src/unwritten_sync > /src/usemem > /src/writemod > +/src/writev_on_pagefault > /src/xfsctl > /src/aio-dio-regress/aio-dio-extend-stat > /src/aio-dio-regress/aio-dio-fcntl-race > diff --git a/src/Makefile b/src/Makefile > index 6b0e4b0..9d54fe7 100644 > --- a/src/Makefile > +++ b/src/Makefile > @@ -9,7 +9,7 @@ TARGETS = dirstress fill fill2 getpagesize holes lstat64 \ > nametest permname randholes runas truncfile usemem \ > mmapcat append_reader append_writer dirperf metaperf \ > devzero feature alloc fault fstest t_access_root \ > - godown resvtest writemod makeextents itrash rename \ > + godown resvtest writemod writev_on_pagefault makeextents itrash rename \ > multi_open_unlink dmiperf unwritten_sync genhashnames t_holes \ > t_mmap_writev t_truncate_cmtime dirhash_collide t_rename_overwrite \ > holetest t_truncate_self t_mmap_dio af_unix t_mmap_stale_pmd \ > diff --git a/src/writev_on_pagefault.c b/src/writev_on_pagefault.c > new file mode 100644 > index 0000000..54845c3 > --- /dev/null > +++ b/src/writev_on_pagefault.c > @@ -0,0 +1,74 @@ > +/* > + * Takes page fault while writev is iterating over the vectors in the IOV > + * > + * Copyright (C) 2017 Red Hat, Inc. All Rights reserved. > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of the GNU General Public License as published by > + * the Free Software Foundation; either version 2 of the License, or > + * (at your option) any later version. > + * > + * This program is distributed in the hope that it will be useful, > + * but WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > + * GNU General Public License for more details. > + * > + * You should have received a copy of the GNU General Public License > + * along with this program; if not, write to the Free Software > + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. > + */ > +#include <stdio.h> > +#include <stdlib.h> > +#include <sys/types.h> > +#include <sys/stat.h> > +#include <fcntl.h> > +#include <string.h> > +#include <unistd.h> > +#include <sys/uio.h> > + > +#define IOV_CNT 3 Make it a default value and can be specified via command line as well? > + > +void > +usage(char *progname) > +{ > + fprintf(stderr, "usage: %s filename\n", progname); > + exit(1); > +} > + > +int main(int argc, char *argv[]) > +{ > + int fd, i; > + size_t ret; > + struct iovec *iov; > + int pagesz = 4096; > + char *data = NULL; > + > + if (argc != 2) > + usage(argv[0]); > + > + pagesz = getpagesize(); > + data = malloc(pagesz * IOV_CNT); Check return value of malloc. > + > + /* no pre-writing on the buffer before writev */ > + > + iov = calloc(IOV_CNT, sizeof(struct iovec)); Check return value of iov. And from this line on, you're mixing tab and spaces for indention. Please use tab consistently. Thanks, Eryu > + for (i = 0; i < IOV_CNT; i++) { > + (iov + i)->iov_base = (data + pagesz * i); > + (iov + i)->iov_len = 1; > + } > + > + if ((fd = open(argv[1], O_TRUNC|O_CREAT|O_RDWR, 0644)) < 0) { > + perror("open failed"); > + return 1; > + } > + > + > + ret = writev(fd, iov, IOV_CNT); > + if (ret < 0) > + perror("writev failed"); > + else > + printf("wrote %d bytes\n", (int)ret); > + > + close(fd); > + return 0; > +} > diff --git a/tests/generic/443 b/tests/generic/443 > new file mode 100755 > index 0000000..89e12b0 > --- /dev/null > +++ b/tests/generic/443 > @@ -0,0 +1,61 @@ > +#! /bin/bash > +# FS QA Test 443 > +# > +# Takes page fault while writev is iterating over the vectors in the IOV > +# > +#----------------------------------------------------------------------- > +# Copyright (c) 2017 Red Hat, Inc. All Rights Reserved. > +# > +# This program is free software; you can redistribute it and/or > +# modify it under the terms of the GNU General Public License as > +# published by the Free Software Foundation. > +# > +# This program is distributed in the hope that it would be useful, > +# but WITHOUT ANY WARRANTY; without even the implied warranty of > +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > +# GNU General Public License for more details. > +# > +# You should have received a copy of the GNU General Public License > +# along with this program; if not, write the Free Software Foundation, > +# Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA > +#----------------------------------------------------------------------- > +# > + > +seq=`basename $0` > +seqres=$RESULT_DIR/$seq > +echo "QA output created by $seq" > + > +here=`pwd` > +tmp=/tmp/$$ > +status=1 # failure is the default! > +trap "_cleanup; exit \$status" 0 1 2 3 15 > + > +_cleanup() > +{ > + cd / > + rm -f $tmp.* > +} > + > +# get standard environment, filters and checks > +. ./common/rc > +. ./common/filter > + > +# remove previous $seqres.full before test > +rm -f $seqres.full > + > +# real QA test starts here > +_supported_fs generic > +_supported_os Linux > +_require_test > +_require_test_program "writev_on_pagefault" > + > +# This program use several vectors for writev(), the kernel goes over them > +# one at a time, copying them from userspace, getting the user data ready > +# for IO. If it takes a page fault while iterating over the vectors in the > +# IOV, it stops, and sends what it got so far. We try to find a bug at this > +# moment. > +$here/src/writev_on_pagefault $TEST_DIR/testfile.$seq > + > +# success, all done > +status=0 > +exit > diff --git a/tests/generic/443.out b/tests/generic/443.out > new file mode 100644 > index 0000000..2d5acf9 > --- /dev/null > +++ b/tests/generic/443.out > @@ -0,0 +1,2 @@ > +QA output created by 443 > +wrote 3 bytes > diff --git a/tests/generic/group b/tests/generic/group > index ab1e9d3..5046c97 100644 > --- a/tests/generic/group > +++ b/tests/generic/group > @@ -443,3 +443,4 @@ > 440 auto quick encrypt > 441 auto quick > 442 blockdev > +443 auto quick rw > -- > 2.7.5 > > -- > To unsubscribe from this list: send the line "unsubscribe fstests" in > the body of a message to majordomo@xxxxxxxxxxxxxxx > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe fstests" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html