Message ID | F3D3F6AC3820BB4C9FCA340DB5C32CB40386FA88@dggeml531-mbx.china.huawei.com |
---|---|
State | Superseded |
Headers | show |
Series | 答复: [PATCH] openposix/fork/11-1.c: Clean up temporary file | expand |
On Fri, Feb 19, 2021 at 5:34 PM zhaogongyi <zhaogongyi@huawei.com> wrote: > Hi Li, > > > > I think we need to delete the tempfile after the testing(in the cleanup > phase). > > > > In this case, we just hope that call unlink after open/mkstemp since > unlink would not destroy the inode, > > > > so call unlink before fork would not affect the test? > Right, but better to put behind of close(fd) that makes logic more clear. > > BTW, seems the child report a wrong status of the locked file, while > > fcntl(fd, F_SETLK, &fl) return -1 and sets errno to EACCES or EAGAIN, > > shouldn't that mean the file already been locked by the parent? > And what do u think about this?
diff --git a/testcases/open_posix_testsuite/conformance/interfaces/fork/11-1.c b/testcases/open_posix_testsuite/conformance/interfaces/fork/11-1.c index a43bc274c..e9b18958c 100644 --- a/testcases/open_posix_testsuite/conformance/interfaces/fork/11-1.c +++ b/testcases/open_posix_testsuite/conformance/interfaces/fork/11-1.c @@ -79,6 +79,8 @@ int main(void) return result; } + unlink(path_template); Why do we need to delete the tempfile before acquiring a lock? Or did you mean doing this after the testing(in the cleanup phase).
Hi Li, I think we need to delete the tempfile after the testing(in the cleanup phase). In this case, we just hope that call unlink after open/mkstemp since unlink would not destroy the inode, so call unlink before fork would not affect the test? 发件人: Li Wang [mailto:liwang@redhat.com] 发送时间: 2021年2月19日 15:45 收件人: zhaogongyi <zhaogongyi@huawei.com> 抄送: LTP List <ltp@lists.linux.it> 主题: Re: [LTP] [PATCH] openposix/fork/11-1.c: Clean up temporary file Hi Gongyi, On Fri, Feb 19, 2021 at 1:05 PM zhao gongyi <zhaogongyi@huawei.com<mailto:zhaogongyi@huawei.com>> wrote: We need to clean up temporary file /tmp/fork-11-1-XXXXXX which created by mkstemp. Signed-off-by: zhao gongyi <zhaogongyi@huawei.com<mailto:zhaogongyi@huawei.com>> --- .../open_posix_testsuite/conformance/interfaces/fork/11-1.c | 2 ++ 1 file changed, 2 insertions(+) BTW, seems the child report a wrong status of the locked file, while fcntl(fd, F_SETLK, &fl) return -1 and sets errno to EACCES or EAGAIN, shouldn't that mean the file already been locked by the parent? + if (fcntl(fd, F_SETLK, &fl) == -1) { printf("Could not set initial lock: %s (%d)\n", strerror(errno), errno); -- 2.17.1 -- Mailing list info: https://lists.linux.it/listinfo/ltp -- Regards, Li Wang