From: Paolo Bonzini <pbonzini@redhat.com>
To: qemu-devel@nongnu.org
Subject: [Qemu-devel] [PATCH 4/4] qemu-nbd: do not start the block layer in the parent
Date: Fri, 28 Oct 2011 12:17:57 +0200 [thread overview]
Message-ID: <1319797077-25441-5-git-send-email-pbonzini@redhat.com> (raw)
In-Reply-To: <1319797077-25441-1-git-send-email-pbonzini@redhat.com>
Forking and threading do not behave very well together. With qemu-nbd in
-c mode it could happen that the thread pool is started in the parent, and
the children (the one actually doing the I/O) is left without working I/O.
Fix this by only running bdrv_init in the child process.
Reported-by: Pierre Riteau <Pierre.Riteau@irisa.fr>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
qemu-nbd.c | 31 ++++++++++++++-----------------
1 files changed, 14 insertions(+), 17 deletions(-)
diff --git a/qemu-nbd.c b/qemu-nbd.c
index 5031158..962952c 100644
--- a/qemu-nbd.c
+++ b/qemu-nbd.c
@@ -371,21 +371,6 @@ int main(int argc, char **argv)
return 0;
}
- bdrv_init();
-
- bs = bdrv_new("hda");
-
- if ((ret = bdrv_open(bs, argv[optind], flags, NULL)) < 0) {
- errno = -ret;
- err(EXIT_FAILURE, "Failed to bdrv_open '%s'", argv[optind]);
- }
-
- fd_size = bs->total_sectors * 512;
-
- if (partition != -1 &&
- find_partition(bs, partition, &dev_offset, &fd_size))
- err(EXIT_FAILURE, "Could not find partition %d", partition);
-
if (device) {
pid_t pid;
int sock;
@@ -418,7 +403,6 @@ int main(int argc, char **argv)
size_t blocksize;
ret = 0;
- bdrv_close(bs);
do {
sock = unix_socket_outgoing(socket);
@@ -473,8 +457,21 @@ int main(int argc, char **argv)
/* children */
}
+ bdrv_init();
+ bs = bdrv_new("hda");
+ if ((ret = bdrv_open(bs, argv[optind], flags, NULL)) < 0) {
+ errno = -ret;
+ err(EXIT_FAILURE, "Failed to bdrv_open '%s'", argv[optind]);
+ }
+
+ fd_size = bs->total_sectors * 512;
+
+ if (partition != -1 &&
+ find_partition(bs, partition, &dev_offset, &fd_size)) {
+ err(EXIT_FAILURE, "Could not find partition %d", partition);
+ }
+
sharing_fds = g_malloc((shared + 1) * sizeof(int));
-
if (socket) {
sharing_fds[0] = unix_socket_incoming(socket);
} else {
--
1.7.6.4
next prev parent reply other threads:[~2011-10-28 10:18 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-10-28 10:17 [Qemu-devel] [PATCH 0/4] fix qemu-nbd -c Paolo Bonzini
2011-10-28 10:17 ` [Qemu-devel] [PATCH 1/4] qemu-nbd: exit if the child exits before a socket connection is established Paolo Bonzini
2011-10-28 10:17 ` [Qemu-devel] [PATCH 2/4] qemu-nbd: include our own err/errx implementation Paolo Bonzini
2011-10-28 10:17 ` [Qemu-devel] [PATCH 3/4] qemu-nbd: report errors to syslog when daemonized Paolo Bonzini
2011-10-28 10:17 ` Paolo Bonzini [this message]
2011-10-28 11:56 ` [Qemu-devel] [PATCH 4/4] qemu-nbd: do not start the block layer in the parent Pierre Riteau
2011-10-28 11:57 ` Paolo Bonzini
2011-10-28 12:16 ` Pierre Riteau
2011-10-28 12:17 ` Paolo Bonzini
2011-11-04 9:46 ` Paolo Bonzini
2011-11-04 10:31 ` Kevin Wolf
2011-11-04 11:10 ` Paolo Bonzini
2011-11-04 11:22 ` Kevin Wolf
2011-11-04 11:25 ` Paolo Bonzini
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1319797077-25441-5-git-send-email-pbonzini@redhat.com \
--to=pbonzini@redhat.com \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).