Messages in this thread | | | Date | Tue, 20 Oct 2015 13:03:32 +0200 | From | Peter Zijlstra <> | Subject | Re: [PATCH 2/2] x86, perf: Optimize stack walk user accesses |
| |
On Mon, Oct 19, 2015 at 03:54:29PM -0700, Andi Kleen wrote: > @@ -2307,7 +2312,13 @@ perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs) > frame.next_frame = NULL; > frame.return_address = 0; > > - bytes = copy_from_user_nmi(&frame, fp, sizeof(frame)); > + if (!access_ok(VERIFY_READ, fp, 16)) > + break; > + > + bytes = __copy_from_user_nmi(&frame.next_frame, fp, 8); > + if (bytes != 0) > + break; > + bytes = __copy_from_user_nmi(&frame.return_address, fp+8, 8); > if (bytes != 0) > break; >
The previous patch that introduces this function states that any caller must have pagefault_disable() or be from interrupt context.
Perf can call this function from !interrupt context (imagine a tracepoint or other software event), should we therefore not add a pagefault_disable()/enable() pair around the entire while() loop?
| |