getsockopt() early argument sanity checking
Quoted 13 lines · Arjan
Quoted 4 lines
We're on UP.
sys_getsockopt()doesget_user()(due to the patch) and makes sure that the passed *optlen is sane. Even if thisget_user()sleeps, the value it returns in "len" is what's currently in memory at the time of theget_user()return (correct?) Then an underlyingQuoted 8 lines
*getsockopt()function does anotherget_user()on optlen (same address), without doing any other user-space data accesses or anything else that could sleep first. Is it possible that this secondget_user()invocation would sleep? I think not since it's the same address that we've just read a value from, we did not leave kernel space, and we're on UP (so no other processor could have changed the mapping). So the patch appears to be sufficient for this special case (which is not unlikely).
Quoted 3 lines
this reasoning goes out the window with kernel preemption of course ;)
Or O_DIRECT? I'm not sure what's easier to time, a kernel preemption or a DMA to the user address.
--
Manfred