-
jclulow
-
jperkinyeh, libcxx stopped building on us in 15.x
-
jperkinus-central.manta.mnx.io/pkgsrc/publ…221204.2250/libcxx-15.0.6/build.log if someone wants to get to it before me
-
tsoome_copec I recommend ipf-howto, it does nice job walking the ipf features
-
sjorgeIs libcxx like a cross platform c++ libc?
-
sjorgeOr am I understanding wrong what it is
-
gitomat[illumos-gate] 15206 setcontext(2) should not restore %fsbase -- Andy Fiddaman <illumos⊙fn>
-
jperkinsjorge: it's llvm's equivalent of gcc's libstdc++
-
jperkinyou can configure clang to use libstdc++ but I've always tried to ensure it uses the full llvm suite
-
jperkinI have libcxx building now, fwiw
-
paulfahh the chequered history of C++ libraries in the Solaris world
-
sjorgeI see
-
sjorgeSo it's rather important that it keeps working for us
-
toasterkingHello, it's Zig guy again. After some recent changes in the project, it seems like my issues building the lang. from source are fixed. The only issue I'm running into right now is uses the linker flag `stack-size=0x10000000`, which isn't implemented in the solaris linker. Is there a workaround or alternative flag I can use?
-
EllenorWill Zig work without it?
-
toasterkingIf I comment it out it dumps core
-
toasterkingThis is a consequence of the bootstrapping project, but basically the compiler is machine translated now, which spits out a huge C file.
-
andyfAre you building the final binary with a map file on illumos?
-
andyfI was looking at the PT_SUNWSTACK segment type, but it looks like it's only the permission bits there that mean anything
-
andyf(which is different to the GNU one, where the size is the desired stack size)
-
jbkhas anyone seen openssh complain about setsockopt(SO_KEEPALIVE)? ISTR something, but can't seem to find a ticket
-
jbk(returning EINVAL)
-
tsoome_yes
-
tsoome_but not recently
-
tsoome_I did try to track it, but did not get too far as it did appear at random times...
-
sjorgei think if the severside keepalive thing is configured
-
sjorgebut bot checked recenrly
-
sjorge*not