On 25 Jan, Mathew Kanner wrote: > On Jan 24, Don Lewis wrote: >> The is a buffer overflow in the interrupt handler in the vchan code that >> occasionally stomps on something on the kernel heap and causes panics in >> other parts of the kernel. I tracked down the cause of the buffer >> overflow with the help of Stefan Ehmann and made a number of changes to >> the sound/pcm/* code to try to avoid triggering the overflow. > >> [...] > > Don, > I think it should be noted that vchans aren't always enabled > by default and those testing this patch should explicitly make sure > that vchans are enabled either via > > sysctl hw.snd.pcmX.vchans=5 That's a good point. I'd also like to make sure that I didn't break anything in the non-vchan case, so testing with vchans on and off would be useful.Received on Sun Jan 25 2004 - 13:20:09 UTC
This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:37:39 UTC