**** BEGIN LOGGING AT Wed Mar 18 02:59:58 2020 Mar 18 10:18:08 mornin Mar 18 16:48:55 khem: what toolchain testing are yocto/oe doing right now for the internal toolchains? Mar 18 16:54:37 kergoth: the gcc test suite is ran Mar 18 16:54:47 can we kill external toolchains yet please Mar 18 16:57:23 that'd be nice, sstate + hostequiv is getting us closer Mar 18 16:58:25 name one good use case for external toolchains Mar 18 16:58:33 apart from "vendor locking" Mar 18 16:59:26 (genuine question, i'm curious) Mar 18 17:00:39 in the past there was a build time benefit beyond the vendor choice and confidence level in the toolchain maintenance. none of which are as much of a deal now Mar 18 17:01:36 vendor is less of an issue considering you can get ex. linaro toolchain in yocto layer form, build time is covered by sstate, and there's definite quality in the maintenance of the yocto toolchain, so i don't see much of a compelling reason anymore Mar 18 17:03:14 i had to deal with an team that were using the linaro toolchain and it was causing all sorts of pain Mar 18 17:03:41 not saying anything about their quality personally, just opens up options for folks who want them Mar 18 17:03:46 i've never used a linaro toolchain Mar 18 17:05:06 kergoth: its running all regression tesuites for binutils/gdb/gcc Mar 18 17:05:10 and glibc Mar 18 19:12:09 external toolchain is one of the paths people use to build CUDA with meta-tegra, I think Mar 18 19:13:17 the other is making older gcc/g++ work by forward porting recipes and hacking things, both are less than ideal Mar 18 22:55:50 I think there must be a section who do it for a shoulder to cry on or more bluntly to get a number to call if something goes wrong Mar 18 22:59:16 khem: heh **** ENDING LOGGING AT Thu Mar 19 03:00:02 2020