Message ID | 1394532550-21857-2-git-send-email-marcel.a@redhat.com |
---|---|
State | New |
Headers | show |
On Tue, Mar 11, 2014 at 12:09:09PM +0200, Marcel Apfelbaum wrote: > @@ -78,12 +79,16 @@ static int socket_accept(int sock) > struct sockaddr_un addr; > socklen_t addrlen; > int ret; > + struct timeval timeout = { .tv_sec = SOCKET_TIMEOUT, > + .tv_usec = 0 }; > + > + setsockopt(sock, SOL_SOCKET, SO_RCVTIMEO, (void *)&timeout, > + sizeof(timeout)); > > addrlen = sizeof(addr); > do { > ret = accept(sock, (struct sockaddr *)&addr, &addrlen); > } while (ret == -1 && errno == EINTR); > - g_assert_no_errno(ret); > close(sock); Did you mean to leave SO_RCVTIMEO set after this function completes? > @@ -91,7 +96,7 @@ static int socket_accept(int sock) > > static void kill_qemu(QTestState *s) > { > - if (s->qemu_pid != -1) { > + if (s && s->qemu_pid != -1) { > kill(s->qemu_pid, SIGTERM); > waitpid(s->qemu_pid, NULL, 0); > } This is a bug in libqtest.c, please don't silence the crash. kill_qemu() gets called from the SIGABRT signal handler but I forgot that global_qtest isn't initialized yet while qtest_init() executes. In other words, the cleanup is broken if we fail inside qtest_init(). Can you drop this hunk and I'll send a patch to fix the underlying issue? > @@ -153,6 +158,8 @@ QTestState *qtest_init(const char *extra_args) > g_free(socket_path); > g_free(qmp_socket_path); > > + g_assert(s->fd >= 0 && s->qmp_fd >= 0); > + We probably shouldn't socket_accept() s->qmp_fd if s->fd already failed. Otherwise we'll wait another 5 seconds for the timeout to explire: s->fd = socket_accept(sock); if (s->fd >= 0) { s->qmp_fd = socket_accept(qmpsock); }
On Tue, 2014-03-11 at 13:40 +0100, Stefan Hajnoczi wrote: > On Tue, Mar 11, 2014 at 12:09:09PM +0200, Marcel Apfelbaum wrote: > > @@ -78,12 +79,16 @@ static int socket_accept(int sock) > > struct sockaddr_un addr; > > socklen_t addrlen; > > int ret; > > + struct timeval timeout = { .tv_sec = SOCKET_TIMEOUT, > > + .tv_usec = 0 }; > > + > > + setsockopt(sock, SOL_SOCKET, SO_RCVTIMEO, (void *)&timeout, > > + sizeof(timeout)); > > > > addrlen = sizeof(addr); > > do { > > ret = accept(sock, (struct sockaddr *)&addr, &addrlen); > > } while (ret == -1 && errno == EINTR); > > - g_assert_no_errno(ret); > > close(sock); > > Did you mean to leave SO_RCVTIMEO set after this function completes? Yes, I don't think it hurts. A 5 sec timeout should be like infinite, Qemu running on the same machine. If you think > > > @@ -91,7 +96,7 @@ static int socket_accept(int sock) > > > > static void kill_qemu(QTestState *s) > > { > > - if (s->qemu_pid != -1) { > > + if (s && s->qemu_pid != -1) { > > kill(s->qemu_pid, SIGTERM); > > waitpid(s->qemu_pid, NULL, 0); > > } > > This is a bug in libqtest.c, please don't silence the crash. I didn't see it like hiding a crash, I thought that if there is any problem during init it is because the Qemu failed to start, meaning that you don't have a process to kill (Qemu exited already). Al of the above happens -> you don't have a global state. Anyway, if you have a better way to deal with it, I have nothing against it :) Thanks, Marcel > > kill_qemu() gets called from the SIGABRT signal handler but I forgot > that global_qtest isn't initialized yet while qtest_init() executes. > > In other words, the cleanup is broken if we fail inside qtest_init(). > Can you drop this hunk and I'll send a patch to fix the underlying > issue? > > > @@ -153,6 +158,8 @@ QTestState *qtest_init(const char *extra_args) > > g_free(socket_path); > > g_free(qmp_socket_path); > > > > + g_assert(s->fd >= 0 && s->qmp_fd >= 0); > > + > > We probably shouldn't socket_accept() s->qmp_fd if s->fd already failed. > Otherwise we'll wait another 5 seconds for the timeout to explire: Yes, I already had this chunk, I have no idea why I dropped it, I'll return it, thanks. Thanks, Marcel > > s->fd = socket_accept(sock); > if (s->fd >= 0) { > s->qmp_fd = socket_accept(qmpsock); > }
On Tue, 2014-03-11 at 14:51 +0200, Marcel Apfelbaum wrote: > On Tue, 2014-03-11 at 13:40 +0100, Stefan Hajnoczi wrote: > > On Tue, Mar 11, 2014 at 12:09:09PM +0200, Marcel Apfelbaum wrote: > > > @@ -78,12 +79,16 @@ static int socket_accept(int sock) > > > struct sockaddr_un addr; > > > socklen_t addrlen; > > > int ret; > > > + struct timeval timeout = { .tv_sec = SOCKET_TIMEOUT, > > > + .tv_usec = 0 }; > > > + > > > + setsockopt(sock, SOL_SOCKET, SO_RCVTIMEO, (void *)&timeout, > > > + sizeof(timeout)); > > > > > > addrlen = sizeof(addr); > > > do { > > > ret = accept(sock, (struct sockaddr *)&addr, &addrlen); > > > } while (ret == -1 && errno == EINTR); > > > - g_assert_no_errno(ret); > > > close(sock); > > > > Did you mean to leave SO_RCVTIMEO set after this function completes? > Yes, I don't think it hurts. A 5 sec timeout should be like infinite, > Qemu running on the same machine. If you think ... otherwise, I can remove the timeout, but I think it is OK. > > > > > > @@ -91,7 +96,7 @@ static int socket_accept(int sock) > > > > > > static void kill_qemu(QTestState *s) > > > { > > > - if (s->qemu_pid != -1) { > > > + if (s && s->qemu_pid != -1) { > > > kill(s->qemu_pid, SIGTERM); > > > waitpid(s->qemu_pid, NULL, 0); > > > } > > > > This is a bug in libqtest.c, please don't silence the crash. > I didn't see it like hiding a crash, I thought that if there > is any problem during init it is because the Qemu failed to start, > meaning that you don't have a process to kill (Qemu exited already). > Al of the above happens -> you don't have a global state. > > Anyway, if you have a better way to deal with it, I have nothing against it :) > > Thanks, > Marcel > > > > > kill_qemu() gets called from the SIGABRT signal handler but I forgot > > that global_qtest isn't initialized yet while qtest_init() executes. > > > > In other words, the cleanup is broken if we fail inside qtest_init(). > > Can you drop this hunk and I'll send a patch to fix the underlying > > issue? I dropped it, please take care of it as it gets a segmentation fault if we abort in qtest_init. Thanks, Marcel > > > > > @@ -153,6 +158,8 @@ QTestState *qtest_init(const char *extra_args) > > > g_free(socket_path); > > > g_free(qmp_socket_path); > > > > > > + g_assert(s->fd >= 0 && s->qmp_fd >= 0); > > > + > > > > We probably shouldn't socket_accept() s->qmp_fd if s->fd already failed. > > Otherwise we'll wait another 5 seconds for the timeout to explire: > Yes, I already had this chunk, I have no idea why I dropped it, I'll > return it, thanks. > > Thanks, > Marcel > > > > s->fd = socket_accept(sock); > > if (s->fd >= 0) { > > s->qmp_fd = socket_accept(qmpsock); > > } > > > >
On Tue, Mar 11, 2014 at 03:04:22PM +0200, Marcel Apfelbaum wrote: > On Tue, 2014-03-11 at 14:51 +0200, Marcel Apfelbaum wrote: > > On Tue, 2014-03-11 at 13:40 +0100, Stefan Hajnoczi wrote: > > > On Tue, Mar 11, 2014 at 12:09:09PM +0200, Marcel Apfelbaum wrote: > > > > @@ -78,12 +79,16 @@ static int socket_accept(int sock) > > > > struct sockaddr_un addr; > > > > socklen_t addrlen; > > > > int ret; > > > > + struct timeval timeout = { .tv_sec = SOCKET_TIMEOUT, > > > > + .tv_usec = 0 }; > > > > + > > > > + setsockopt(sock, SOL_SOCKET, SO_RCVTIMEO, (void *)&timeout, > > > > + sizeof(timeout)); > > > > > > > > addrlen = sizeof(addr); > > > > do { > > > > ret = accept(sock, (struct sockaddr *)&addr, &addrlen); > > > > } while (ret == -1 && errno == EINTR); > > > > - g_assert_no_errno(ret); > > > > close(sock); > > > > > > Did you mean to leave SO_RCVTIMEO set after this function completes? > > Yes, I don't think it hurts. A 5 sec timeout should be like infinite, > > Qemu running on the same machine. If you think > ... otherwise, I can remove the timeout, but I think it is OK. I think you are right. I checked that the qtest protocol has no long-running operations. It doesn't seem realistic that any qtest command would take 5 seconds or longer. So let's leave in the timeout. Stefan
diff --git a/tests/libqtest.c b/tests/libqtest.c index f587d36..f1ba254 100644 --- a/tests/libqtest.c +++ b/tests/libqtest.c @@ -34,6 +34,7 @@ #include "qapi/qmp/json-parser.h" #define MAX_IRQ 256 +#define SOCKET_TIMEOUT 5 QTestState *global_qtest; @@ -78,12 +79,16 @@ static int socket_accept(int sock) struct sockaddr_un addr; socklen_t addrlen; int ret; + struct timeval timeout = { .tv_sec = SOCKET_TIMEOUT, + .tv_usec = 0 }; + + setsockopt(sock, SOL_SOCKET, SO_RCVTIMEO, (void *)&timeout, + sizeof(timeout)); addrlen = sizeof(addr); do { ret = accept(sock, (struct sockaddr *)&addr, &addrlen); } while (ret == -1 && errno == EINTR); - g_assert_no_errno(ret); close(sock); return ret; @@ -91,7 +96,7 @@ static int socket_accept(int sock) static void kill_qemu(QTestState *s) { - if (s->qemu_pid != -1) { + if (s && s->qemu_pid != -1) { kill(s->qemu_pid, SIGTERM); waitpid(s->qemu_pid, NULL, 0); } @@ -153,6 +158,8 @@ QTestState *qtest_init(const char *extra_args) g_free(socket_path); g_free(qmp_socket_path); + g_assert(s->fd >= 0 && s->qmp_fd >= 0); + s->rx = g_string_new(""); for (i = 0; i < MAX_IRQ; i++) { s->irq_level[i] = false;
'socket_accept' waits for Qemu to init its unix socket. If Qemu encounters an error during command line parsing, it can exit before initializing the communication channel. Using a timeout for sockets fixes the issue. Signed-off-by: Marcel Apfelbaum <marcel.a@redhat.com> --- tests/libqtest.c | 11 +++++++++-- 1 file changed, 9 insertions(+), 2 deletions(-)