[PATCH] Fix kobject uevent string handling errors (updated)

2007-08-27 Thread Mathieu Desnoyers
Hi Andrew, I thought you were going to use Kay Sievers's patch, which adds error value checking to much of the add_event_var callers, so I did not bother sending the updated version of my patch. Here it is, with the changelog. --- Fix kobject uevent string handling errors - increment env->bufle

Re: [PATCH] Fix kobject uevent string handling errors

2007-08-25 Thread Mathieu Desnoyers
* Kay Sievers ([EMAIL PROTECTED]) wrote: > env->envp[env->envp_idx++] = &env->buf[env->buflen]; > > env->buflen += len + 1; > > Index: linux-2.6-lttng/drivers/firmware/dmi-id.c > > === > > --- linux-2.6-lttng.orig/drivers/fir

Re: [PATCH] Fix kobject uevent string handling errors

2007-08-24 Thread Greg KH
On Sat, Aug 25, 2007 at 12:17:54AM -0400, Mathieu Desnoyers wrote: > Fix kobject uevent string handling errors > > - increment env->buflen in dmi-id.c > - fix off-by-one in add_uevent_var in error checking of vsnprintf > - add warnings when add_uevent_var. Proper handling of its return values > s

[PATCH] Fix kobject uevent string handling errors

2007-08-24 Thread Mathieu Desnoyers
Fix kobject uevent string handling errors - increment env->buflen in dmi-id.c - fix off-by-one in add_uevent_var in error checking of vsnprintf - add warnings when add_uevent_var. Proper handling of its return values should really be done by the callers, but they aren't, so things currently fa