On Wed, Jul 29, 2009 at 10:56:48PM -0400, Aron Griffis wrote:
Matthias Bolte wrote: [Wed Jul 29 2009, 08:58:18PM EDT]
2009/7/29 Daniel Veillard <veillard@redhat.com>:
On Wed, Jul 29, 2009 at 08:05:37PM +0200, Daniel Veillard wrote:
On Wed, Jul 29, 2009 at 12:52:30PM -0400, Aron Griffis wrote:
This is a resend of take 2 to fix formatting problems in the patch. No other changes.
As far as I can tell, there's no reason to format the device string in brAddTap(). Delegate the job to TUNSETIFF, thereby removing the loop and the MAX_TAP_ID artificial limit. This patch allows me to get 421 guests running before hitting other limits.
haha ! that one worked :-)
I will review and apply, thanks !
Actually just looking at brAddTap() after patching makes it clear, and based on Mark and Dan feedback great ! Applied and commited to git, thanks !
Daniel
This patch breaks -Werror, because GCC is unhappy with the initializer for ifreq.
Sorry I should have caught that
Maybe instead of the nested curlies, we should revert to memset as it was doing previously:
diff --git a/src/bridge.c b/src/bridge.c index ec37192..58e480e 100644 --- a/src/bridge.c +++ b/src/bridge.c @@ -465,7 +465,7 @@ brAddTap(brControl *ctl, int *tapfd) { int fd, len; - struct ifreq ifr = {0}; + struct ifreq ifr;
if (!ctl || !ctl->fd || !bridge || !ifname) return EINVAL; @@ -473,6 +473,8 @@ brAddTap(brControl *ctl, if ((fd = open("/dev/net/tun", O_RDWR)) < 0) return errno;
+ memset(&ifr, 0, sizeof(ifr)); + ifr.ifr_flags = IFF_TAP|IFF_NO_PI;
yeah the memset was the status quo before, let's go back to using this. Applied and commited, thanks ! Daniel -- Daniel Veillard | libxml Gnome XML XSLT toolkit http://xmlsoft.org/ daniel@veillard.com | Rpmfind RPM search engine http://rpmfind.net/ http://veillard.com/ | virtualization library http://libvirt.org/