- User Since
- May 20 2019, 12:58 AM (104 w, 1 d)
Oct 8 2019
Without the -debug parameter the test can be false positive as it won't reach the assert in the PDB code (in addLinkerModuleCoffGroup) then
but I'd like to understand what is going on. Are you at the repo head?
The output on my Arch Linux machine matches the strings the test checks for - is it normal for these values to change? Should we ignore them?
I updated the patch to add --match-full-lines as requested. Can you commit the patch for me, as I don't have commit access?
So, you are creating an executable that has an import table that has only ordinals. Is my understanding correct? Indeed, that condition is rare and I've thought of that case before when I implemented this part.
Yes, that's what we were having issues with.
Oct 7 2019
I updated the patch and included a test case. I ran the test case before and after the change to make sure it behaves and tests correctly.
Oct 6 2019
Thanks for taking the time to give me some suggestions.
Oct 3 2019
Oct 2 2019
To be honest, I know almost nothing about PDB. I'm only changing this because I traced my own build problems to it (Xbox-specific, we're creating xbe-files from relocatable PE files including DWARF-4 debug info generated by clang - log can be found here.
This patch resolves the issues we've encountered when building on platforms that updated to LLD 9, but I don't know whether this is an appropriate (or correct) solution. I also don't know how I would construct a test case for this.
Aug 31 2019
Thanks for accepting! Can you commit the patch for me? I don't have commit access myself.
Aug 29 2019
May 21 2019
I updated the patch to include a power-of-two check.
May 20 2019
Changed Config member name, added a test.