| Message ID | 1493759813-15786-1-git-send-email-gorcunov@virtuozzo.com |
|---|---|
| State | Rejected |
| Series | "util: Don't swear on interrupted close syscall" |
| Headers | show |
diff --git a/criu/util.c b/criu/util.c index cb1e8fd..f5ed388 100644 --- a/criu/util.c +++ b/criu/util.c @@ -183,8 +183,12 @@ int close_safe(int *fd) ret = close(*fd); if (!ret) *fd = -1; - else - pr_perror("Unable to close fd %d", *fd); + else { + if (errno != EINTR) + pr_perror("Unable to close fd %d", *fd); + else + ret = 0, *fd = -1; + } } return ret;
On 05/03/2017 12:16 AM, Cyrill Gorcunov wrote: > In case of EINTR here the descriptor will be > closed anyway. We do we have this EINTR here at all? > http://lkml.indiana.edu/hypermail/linux/kernel/0509.1/0877.html > > Signed-off-by: Cyrill Gorcunov <gorcunov@virtuozzo.com> > --- > criu/util.c | 8 ++++++-- > 1 file changed, 6 insertions(+), 2 deletions(-) > > diff --git a/criu/util.c b/criu/util.c > index cb1e8fd..f5ed388 100644 > --- a/criu/util.c > +++ b/criu/util.c > @@ -183,8 +183,12 @@ int close_safe(int *fd) > ret = close(*fd); > if (!ret) > *fd = -1; > - else > - pr_perror("Unable to close fd %d", *fd); > + else { > + if (errno != EINTR) > + pr_perror("Unable to close fd %d", *fd); > + else > + ret = 0, *fd = -1; > + } > } > > return ret; >
On Thu, May 04, 2017 at 04:33:56PM +0300, Pavel Emelyanov wrote: > On 05/03/2017 12:16 AM, Cyrill Gorcunov wrote: > > In case of EINTR here the descriptor will be > > closed anyway. > > We do we have this EINTR here at all? Any syscall may exit with EINTR, except in case of close it is special because file will be closed anyway.
On Thu, May 04, 2017 at 04:36:24PM +0300, Cyrill Gorcunov wrote: > On Thu, May 04, 2017 at 04:33:56PM +0300, Pavel Emelyanov wrote: > > On 05/03/2017 12:16 AM, Cyrill Gorcunov wrote: > > > In case of EINTR here the descriptor will be > > > closed anyway. > > > > We do we have this EINTR here at all? > > Any syscall may exit with EINTR, except in case of close it is > special because file will be closed anyway. But there should be a reason for this EINTR. Do you have a log when close returns EINTR in criu? > _______________________________________________ > CRIU mailing list > CRIU@openvz.org > https://lists.openvz.org/mailman/listinfo/criu
On Fri, May 05, 2017 at 10:51:49AM -0700, Andrei Vagin wrote: > On Thu, May 04, 2017 at 04:36:24PM +0300, Cyrill Gorcunov wrote: > > On Thu, May 04, 2017 at 04:33:56PM +0300, Pavel Emelyanov wrote: > > > On 05/03/2017 12:16 AM, Cyrill Gorcunov wrote: > > > > In case of EINTR here the descriptor will be > > > > closed anyway. > > > > > > We do we have this EINTR here at all? > > > > Any syscall may exit with EINTR, except in case of close it is > > special because file will be closed anyway. > > But there should be a reason for this EINTR. Do you have a log when > close returns EINTR in criu? No, i don't have log.
On Fri, May 05, 2017 at 09:06:53PM +0300, Cyrill Gorcunov wrote: > On Fri, May 05, 2017 at 10:51:49AM -0700, Andrei Vagin wrote: > > On Thu, May 04, 2017 at 04:36:24PM +0300, Cyrill Gorcunov wrote: > > > On Thu, May 04, 2017 at 04:33:56PM +0300, Pavel Emelyanov wrote: > > > > On 05/03/2017 12:16 AM, Cyrill Gorcunov wrote: > > > > > In case of EINTR here the descriptor will be > > > > > closed anyway. > > > > > > > > We do we have this EINTR here at all? > > > > > > Any syscall may exit with EINTR, except in case of close it is > > > special because file will be closed anyway. > > > > But there should be a reason for this EINTR. Do you have a log when > > close returns EINTR in criu? > > No, i don't have log. Why did you decide to make this patch?;)
On Fri, May 05, 2017 at 11:38:51AM -0700, Andrei Vagin wrote: > > No, i don't have log. > > Why did you decide to make this patch?;) Doesn't matter, drop it please.
In case of EINTR here the descriptor will be closed anyway. http://lkml.indiana.edu/hypermail/linux/kernel/0509.1/0877.html Signed-off-by: Cyrill Gorcunov <gorcunov@virtuozzo.com> --- criu/util.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)