[llvm-commits] [PATCH] [Review request] Adding getLineInforForAddressRange() to DIContext

Eric Christopher echristo at gmail.com
Thu Jan 24 16:10:49 PST 2013


Hi Andy,


> This code is a bit tricky to test because the changes don’t modify the
> behavior of any of the standard llvm tools.  I have been testing in with
> the llvm-jitlistner tool, but that only gets built if Intel JIT profiling
> is enabled in the build, which it isn’t for any of the buildbots.
>

Right.


> Is there a better way to do this?
>
>

Maybe using llvm-rtdyld to load a pre-generated .o file where you know an
address and then use that to look for something? You'd have to add an
option to llvm-rtdyld to print some line information/debug information for
a given .o file but it'd probably be a good way to test?

-eric


> -Andy
>
> ** **
>
> *From:* Eric Christopher [mailto:echristo at gmail.com]
> *Sent:* Thursday, January 24, 2013 1:03 AM
> *To:* Kaylor, Andrew
> *Cc:* llvm-commits
> *Subject:* Re: [PATCH] [Review request] Adding
> getLineInforForAddressRange() to DIContext****
>
> ** **
>
> This looks fine in general. I'd like some testcases though please.****
>
> ** **
>
> Thanks!****
>
> ** **
>
> -eric****
>
> ** **
>
> On Mon, Jan 21, 2013 at 3:37 PM, Kaylor, Andrew <andrew.kaylor at intel.com>
> wrote:****
>
> This patch adds a new method, getLineInfoForAddressRange, to the DIContext
> interface class and provides an implementation for that method in the
> DebugInfo code.****
>
>  ****
>
> This is a subset of a patch I submitted last week (
> http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20130114/162210.html).
> There is nothing new in this patch, I have simply isolated the
> DebugInfo-specific changes to make review easier.****
>
>  ****
>
> The new method will be used by the IntelJITEventListener to provide line
> information for profiling code generated by MCJIT, but these changes are
> not dependent on the JIT event listener changes.  I’ll resubmit the JIT
> event listener changes momentarily for context.****
>
>  ****
>
> -Andy****
>
> ** **
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20130124/ea172c03/attachment.html>


More information about the llvm-commits mailing list