|
From: Bart V. A. <bar...@gm...> - 2006-09-03 18:39:48
|
Being able to record a call stack associated with each load and store would be great. This would allow to report much more precise information about data races. E.g. the ThreadChecker that is part of Intel VTune reports the complete call stack for both first and second accesses involved in a data race. The granularity of this information is one source line: an overview of all data races is displayed in VTune's GUI, and when selecting a data race the source file corresponding to the uppermost stack frame is displayed, and the cursor is positioned on the right source line. On 9/2/06, Julian Seward <js...@ac...> wrote: > > > There's still a way to go, however. I've been thinking about why > drd is difficult to use, what can be done to improve it. [ ... ] > So here's a different impractical idea :-) When drd is recording > races, don't just record the set of read/written addresses, but > something else too: for each written address in the segment, also > record the program counter of the first (or any) writer of that address. > |