There might be a genuine issue, but it's impossible for me to debug this since mods break terms of service. I will need a different way to reproduce it.
FWIW I don't see anything wrong with the commits in question and haven't seen any issues in a fairly large number of games, including FFXIV.
It just crashes my game directly.
Did you copy the appropriate (64-bit) dxgitrace.dll, dxgi.dll and d3d11.dll to the game directory? If so, and if that doesn't work, there's not much you can do.
So, this is interesting @doitsujin.
Still trying to get a proper apitrace, but I have other - possibly related - news.
After commit a6483e02fe28b5472b237e6fe80fd9226da05378 it now seems to be working fine, again.
I'll be testing a bit more and get an apitrace if it's not resolved.
Alright, after multiple compilations and testing it seems like dbc14fe65cf36e2258361aa8c25e6219ebff76ad is the one that fixed the issue.
Somehow the game needs to be started twice (only the first time) before functioning properly after crashing with an earlier build.
Well that's weird. Are you building with -march=native by any chance, or any additional optimizations?
None of these commits should affect correctness on x86, although a6483e02fe28b5472b237e6fe80fd9226da05378 fixes a potential performance problem.
Just meson --cross-file build-win64.txt --buildtype release.
Would a apitrace still be wanted even though it seems fixed as of dbc14fe?
if it's possible to provide one, sure, if not, then.. well, it wouldn't work anyway.
d3d11.dllx1 2019-11dxgi.dllx1 2019-11dxgitrace.dllx1 2019-11
Hi!
Certain mods for FFXIV crash with DXVK after commit d96c22b / 0924bb4.
Bisecting revealed this.
While I realize this is not an officially supported use case, it's still seems to be DXVK regression.
Software information
Final Fantasy XIV - Modded
System information
Apitrace file(s)
Log files