Description of problem: From: Matt Domsch <Matt_Domsch> To: netdev.com If dev->dev_addr is zero, then the memcpy() never takes place, and the same data that was in the caller's buffer is still in the caller's buffer on successful return. The caller can't know that the data in its buffer isn't the right answer. So, if dev->dev_dev_addr == 0, clear the buffer before returning success. Thanks, Matt Version-Release number of selected component (if applicable): 2.6.9-rc1 and earlier 2.6 variants
Created attachment 105967 [details] net-core-dev.c.patch Patch as submitted to netdev.com for 2.6
Created attachment 106049 [details] net-core-dev.c-2.6.patch Updated patch based on discussion on netdev.com. For RHEL4, Red Hat may wish not to include the printk, as there are ~30 applications which will cause that printk to trigger, none of which is really broken.
fixed in cvs
FWIW, DaveM applied the first of my patches, not the latter one. Which got applied to CVS please?
the one from commment #3 (minus the printk)
Verified, patch from #3 minus the printk statement included in kernel 2.6.9-1.751_EL. Spoke to Matt and he stated that he didn't care much for the printk statement in the first place. Closing the issue as this patch meets our needs.