This looks very suspicious, just looking at CL subject lines: 149c1fc2d configs: Remove default kernel value from system creation
On Tue, Apr 30, 2019 at 3:50 PM Gabe Black <[email protected]> wrote: > To bound finding the point things broke, it looks like this change > works/doesn't crash: > > 2a28a4f8d2ac1471ef3ba943fa5f52a2331759bf > > and this one doesn't work/does crash: > > b2efb725921c9387852f439ad20964c428072dd7 > > which leaves about 45 CLs to look through. > > On Tue, Apr 30, 2019 at 3:39 PM Gabe Black <[email protected]> wrote: > >> Oh, also, the reason the kernel symbol table is empty is because the >> "kernel" parameter is showing up as "" in the create() function (at least, >> maybe earlier) for LinuxAlphaSystemParams, even though it's clearly not "" >> in the m5out/config.ini file. I don't know why there's that mismatch. >> >> Gabe >> >> On Tue, Apr 30, 2019 at 3:37 PM Gabe Black <[email protected]> wrote: >> >>> Hey folks, I just noticed that the Alpha regression tests are failing, >>> and it seems to be because the kernel symbol table is empty for some >>> reason. I don't have a lot of time today, but I'll try to find where they >>> stopped working. If you're bored, please feel free to dive in and help >>> figure out what's going on. This may affect more than just Alpha. >>> >>> Gabe >>> >> _______________________________________________ gem5-dev mailing list [email protected] http://m5sim.org/mailman/listinfo/gem5-dev
