search for: inilin

...le -- are there any practical reason for doing this? > Yes, its complexity which is better addressed by another layer. Every piece of complexity, especially in something like the inliner, makes it harder to change and reason about. I'm trying to minimize and isolate the complex parts of the iniliner in preparation to porting it to the new pass manager. This seemed like low-hanging fruit. > you mentioned information loss (aliasing, sanitizer etc) -- are there any > real issues caused by the merging (with tracking bugs)? > I don't have any, but I also don't think we should...

...But today when i tried to boot a DomU with birdge enabled it didnt work giving an error .... Error: Device 0 (vif) could not be connected. Hotplug scripts not working. After that I checked my var/log/messages and It gave lots of xenstore-write segfault,exact error is attached iniline. Jul 12 07:26:46 centosxcat2 kernel: xenstore-read[16643]: segfault at 0000000000000000 rip 0000000000000000 rsp 00007fff2036def8 error 4 Now I dont know whether segfault error were caused by first error or they are independent , I think they occured after i tried to create d...

Chris added alloca merging in the inliner a looooong time ago, 2009. The reason he added it was because at the time we didn't do stack coloring and without it we had serious stack size problems in LLVM. Since then, a few rather important things have changed: - We have reasonably powerful stack coloring support in the backend based on lifetime markers - Clang (the primary frontend I'm