search for: dsutherland

...hat the NDAs have finally expired. Sadly, you can't express the resulting source line information in the debug directives used by any commonly available debugger (that I am aware of). So -- at the very most optimistic -- this approach won't get you anything any time soon. Dean Sutherland dsutherland at cert.org<mailto:dsutherland at cert.org> On Jun 7, 2013, at 12:38 PM, Renato Golin <renato.golin at linaro.org<mailto:renato.golin at linaro.org>> wrote: On 7 June 2013 13:53, Dallman, John <john.dallman at siemens.com<mailto:john.dallman at siemens.com>> wrote: I...

On Tue, Oct 15, 2013 at 12:01 PM, Dean Sutherland <dsutherland at cert.org>wrote: > Here at CERT we've written some prototype tools that use YAML files to > hold a minimal subset of the Clang parse tree. We then combine these files > to perform cross-TU static analysis. We write out *only* the minimal > information required for the partic...

...Of course, that's all a hack-around to enable the cross-TU analysis... We'd be rather disappointed if the YAML library suddenly (a) started supporting a limited set of output file kinds and (b) failed to provide a simple mechanism to add custom file kinds to that set. Dean F. Sutherland dsutherland at cert.org<mailto:dsutherland at cert.org> On Oct 15, 2013, at 11:32 AM, Sean Silva <chisophugis at gmail.com<mailto:chisophugis at gmail.com>> wrote: On Mon, Oct 14, 2013 at 11:33 PM, Shankar Easwaran <shankare at codeaurora.org<mailto:shankare at codeaurora.org>&...

On Mon, Oct 14, 2013 at 11:33 PM, Shankar Easwaran <shankare at codeaurora.org>wrote: > On 10/14/2013 8:20 PM, Sean Silva wrote: > >> On Mon, Oct 14, 2013 at 8:41 PM, Michael Spencer <bigcheesegs at gmail.com >> >wrote: >> >> On Wed, Oct 9, 2013 at 11:23 AM, Shankar Easwaran < >>> shankare at codeaurora.org >>> >>>>

On 7 June 2013 13:53, Dallman, John <john.dallman at siemens.com> wrote: > It needs to be possible to debug code at any optimisation level. > Yes, I agree. But after O1, sequential execution is a big impediment for optimizations, and keeping the debug information valid after so many transformations might pose a big penalty on the passes (time & memory). That was the whole idea

I'm not a LLVM or Clang developer, but I do spend a lot of time teasing software into working with the highest possible optimisation levels where it still works correctly. These guidelines are pretty good, but there are a few details worth considering. It needs to be possible to debug code at any optimisation level. It's acceptable for that to be harder at high optimisation levels, but

...information about the Secure Coding group at CERT; this position (or positions) would fall largely within the Development Tools and Libraries area. I'm happy to provide additional information, but please note that any actual applications must go through the CERT jobs page. Dean F. Sutherland dsutherland at cert.org

On Fri, Nov 8, 2013 at 8:44 AM, Filip Pizlo <fpizlo at apple.com> wrote: > Is the expectation that to utilize this metadata an optimization pass > would have to inspect the body of @f and reason about its behavior given > <args>? > > > Yes. > > > If so, then I think this is pretty bad. If we ever want to parallelize > function passes, then they can't

> On Nov 8, 2013, at 1:13 AM, Chandler Carruth <chandlerc at google.com> wrote: > > >> On Thu, Nov 7, 2013 at 9:39 PM, Filip Pizlo <fpizlo at apple.com> wrote: >> NEW PROPOSAL >> >> The solution is to introduce meta-data that is explicit about how the safe-to-execute condition ought to be evaluated. Instead of an SSA use, we can have meta-data that