Re: [PATCH] tools/xenstore/xenstore_core.c

[Rusty Russell <rusty@rustcorp.com.au>]

On Thu, 2005-08-11 at 23:02 +0100, Keir Fraser wrote:
> This one's a bug, but I don't think that's the right fix. Rusty?

Yeah, we used to close stdout earlier, so this used a temporary dup'ed
fd.  It's now much simpler: nice spotting Jerone.

Rusty.

# HG changeset patch
# User Rusty Russell <rusty@rustcorp.com.au>
# Node ID 1f532dacdc4870c235d57c55b2f6051b9ee40419
# Parent  822566f5cfddce3b641b5a1a119dfd592ec69158
Fix --output-pid (broken when we fixed --pid-file).
Thanks to Jerone Young for spotting this.

Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>

diff -r 822566f5cfdd -r 1f532dacdc48 tools/xenstore/xenstored_core.c
--- a/tools/xenstore/xenstored_core.c	Thu Aug 11 22:18:27 2005
+++ b/tools/xenstore/xenstored_core.c	Fri Aug 12 03:08:56 2005
@@ -1587,7 +1587,7 @@
 
 int main(int argc, char *argv[])
 {
-	int opt, *sock, *ro_sock, event_fd, max, tmpout;
+	int opt, *sock, *ro_sock, event_fd, max;
 	struct sockaddr_un addr;
 	fd_set inset, outset;
 	bool dofork = true;
@@ -1673,9 +1673,8 @@
 	restore_existing_connections();
 
 	if (outputpid) {
-		char buffer[20];
-		sprintf(buffer, "%i\n", getpid());
-		write(tmpout, buffer, strlen(buffer));
+		printf("%i\n", getpid());
+		fflush(stdout);
 	}
 
 	/* close stdin/stdout now we're ready to accept connections */

-- 
A bad analogy is like a leaky screwdriver -- Richard Braakman