<noam>
But... what dependency is there?? a.next depends on the *address* of b, which depends only on knowing b's *size*, which dependso nly on type info
<noam>
The value of b.next has no effect on a
<g-w1>
thats why its a bug i think :P
<noam>
Is that just a false dependency detection?
<g-w1>
yea
andreao7 has joined #zig
andreao has quit [Ping timeout: 240 seconds]
andreao7 is now known as andreao
andreao6 has joined #zig
andreao has quit [Ping timeout: 252 seconds]
andreao6 is now known as andreao
artmb has joined #zig
artmb has quit [Ping timeout: 246 seconds]
ur5us has quit [Ping timeout: 258 seconds]
marler8997 has joined #zig
<marler8997>
how would I set the eval branch quota when I'm initializing a ComptimeStringMap?
<andrewrk>
if you remove the range case, it becomes very compact
<ifreund>
andrewrk: are you planning on doing the switch validation functions in sema? I might give them a go if not
<ifreund>
doesn't seem like there's a whole lot left to do on the branch in general
<andrewrk>
that was going to be my next task, but I'm about to break for lunch. you want to go for it? :)
<andrewrk>
Module.SrcLoc.byteOffset needs some work too
<andrewrk>
also last I checked the cbe .h file test cases were failing for some reason
<g-w1>
andrewrk: do you know why setevalbranchquota test is being reported failing on llvm12 tests?
<ifreund>
hmm, maybe I'll look into the cbe test failures instead, I haven't done anything in sema yet
<andrewrk>
g-w1, nope have not looked into it yet
<andrewrk>
when was the last time you checked? I think I resolved an issue related to that a few days ago
<g-w1>
the new pass manager pr mentioned it
<g-w1>
havent compiled llvm12 yet :)
* andrewrk
breaks for lunch
<mq32>
heya o/
<mq32>
andrewrk, have a good one!
<mq32>
how does zig std handle ioctl codes?
<mq32>
i want to create portable zig code and i assume that ioctl codes differ between platforms?
mq32 is now known as ikskuh
aerona has joined #zig
<ifreund>
ikskuh: I don't think it really does yet, but it should probably be done the same way as other info that differs between platforms/architectures
<ifreund>
for now you can just @cImport() sys/ioctl.h
frett27 has quit [Ping timeout: 246 seconds]
<ifreund>
for example we do have SIOCGIFINDEX defined in std/os/bits/linux.zig
<ikskuh>
ifreund: hm. i'd love to have no C deps in that project :/
<ifreund>
ikskuh: You shouldn't need to link libc to include that header
<ikskuh>
hm, i'll see :)
<ifreund>
though you may manually need to tell zig where to look if you don't
notzmv has quit [Read error: Connection reset by peer]
Guest89795 has joined #zig
Guest89795 is now known as notzmv
<andrewrk>
finally, got switch expressions passing tests in zir-memory-layout
<andrewrk>
might even get the branch done today
<andrewrk>
ifreund, you got anything WIP for me to watch out for?
artmb has joined #zig
artmb has quit [Ping timeout: 246 seconds]
ur5us has quit [Ping timeout: 258 seconds]
<g-w1>
seems like emit-h needs some work?
<g-w1>
seems like it is literally trying to return void. ill look into why retvoid is not being used :1:8: error: TODO: C backend: implement value Void