15:14:28 Is there anyone with Linux experience here who can tell me why I can't seem to have 15:14:39 F_OFD_* fcntls defined? 15:14:49 e.g. 15:14:50 util.c:155:14: error: 'F_OFD_SETLK' undeclared (first use in this function); did you mean 'F_SETLK'? 15:17:30 -D__USE_GNU? 15:18:36 I did that in a header file but maybe I need it on the gcc cmdline? 15:19:11 Doesn't help @denk 15:19:25 gcc12 on Debian 12 if this helps... 15:19:29 And yes I've installed the header files. 15:21:02 probably it needs something else, it's always a pain when I try to find answers to such questions 15:21:42 from time to time I just use gcc -E 15:35:03 danmcd: -D_GNU_SOURCE 15:35:42 THANK YOU! 15:35:45 https://paste.dilos.org/?3b4245f13ff5593c#psxp8ZCAXe/mdHbVA8qGKOVjBvcpuJf1Y+ojLcz4hkM= 19:56:55 jperkin: your gcc __STDC_VERSION__ stuff would be good to float around the developer list 19:57:16 I think we've all looked at it and gone "Oh, huh?" before 19:59:22 yeh I have a few things that I'm testing that can likely be unified, like using the dragonfly locale backend 20:24:44 the dragonfly locale thing makes sense -- I think Garrett took the code from an equivalently aged FreeBSD -- but I suspect what it _actually_ means is the locale code in libc needs fixed 20:25:07 probably worth investigating how much it would suck if we then fixed and had to change back 20:25:37 I am semi-familiar with that stuff if there's something specific that can help. 20:27:37 which bit? 20:28:10 Fixing locale code in libc. 20:40:01 I have some patches in review (that I really should finish) that affect how we find the current locale. 20:40:16 but they should be largely orthogonal to what you do once you find it.. 20:45:11 I have heard complaints about C++ locale stuff for a decade, but none that actually explained it 20:45:18 just were like "and there's the locale stuff..." 20:46:02 I think we added the _l stuff initially because it was required there, and then apparently we didn't get it right enough 20:47:13 to be honest, I think the intersection of C++ runtime and OS is something that requires far more from the C++ side than the OS 20:47:20 knowledge-wise 21:42:19 tsoome: Why do we do this ... https://src.illumos.org/source/xref/illumos-gate/usr/src/boot/efi/loader/copy.c?r=f334afcf&fi=efi_loadaddr#190-191 21:43:27 viz., why do we constrain ourselves to memory under 4GB 21:44:05 I have found a system with 507MB at 000000100000 and then nothing really until 62.9GB up at 000100000000 21:47:41 There is a comment that says ""reportedly some 64bit systems are reported to have issues with memory above 4GB" but not really ... any detail on what those issues might be 22:02:14 anyway, https://www.illumos.org/issues/16572 22:02:16 → BUG 16572: loader should be willing to put modules above 4GB if necessary (New)