Re: CVS commit: src/sys/rump
On 09.03.2020 15:31, Taylor R Campbell wrote: >> Module Name:src >> Committed By: kamil >> Date: Mon Mar 9 00:03:00 UTC 2020 >> >> Modified Files: >> src/sys/rump: Makefile.rump >> >> Log Message: >> Build RUMP with -fno-delete-null-pointer-checks on all compilers > > I asked you to hold off on making this change, ten hours before you > proceeded to make it: > > https://mail-index.NetBSD.org/tech-kern/2020/03/08/msg026125.html > > It remains unclear whether the change is actually necessary. > It is necessary. The tool detects real UB and the tool is not broken. I went that far that I summoned people from compilers and the C + C++ standard to confirm this. > Please back this out until the discussion has reached a conclusion, > and please don't unilaterally move ahead to make changes that are > still under active discussion -- especially when someone in the > discussion you initiated has specifically asked you to wait. > > If there is a dispute that has deadlocked in public discussion and > requires a resolution, then you can present a case to core and ask for > a ruling. But right now the discussion is not deadlocked and nobody > has asked for core to step in to resolve any dispute in order to make > progress. > I will revert it for the time being on this demand. signature.asc Description: OpenPGP digital signature
Re: CVS commit: src/sys/rump
> Module Name:src > Committed By: kamil > Date: Mon Mar 9 00:03:00 UTC 2020 > > Modified Files: > src/sys/rump: Makefile.rump > > Log Message: > Build RUMP with -fno-delete-null-pointer-checks on all compilers I asked you to hold off on making this change, ten hours before you proceeded to make it: https://mail-index.NetBSD.org/tech-kern/2020/03/08/msg026125.html It remains unclear whether the change is actually necessary. Please back this out until the discussion has reached a conclusion, and please don't unilaterally move ahead to make changes that are still under active discussion -- especially when someone in the discussion you initiated has specifically asked you to wait. If there is a dispute that has deadlocked in public discussion and requires a resolution, then you can present a case to core and ask for a ruling. But right now the discussion is not deadlocked and nobody has asked for core to step in to resolve any dispute in order to make progress.
Re: CVS commit: [netbsd-9] src/sys
On Mon, Mar 09, 2020 at 02:22:26PM +0100, Tobias Nygren wrote: > /usr/src/sys/arch/amd64/amd64/trap.c: In function 'trap': > /usr/src/sys/arch/amd64/amd64/trap.c:360:2: error: expected expression before > '}' token > > 1.107 christos 356: > 1.121.2.1! martin 357: > MODULE_HOOK_CALL(amd64_oosyscall_hook, (p, frame), > ! 358:ENOSYS, hook_ret); > ! 359:if (hook_ret == 0) > 1.107 christos 360:} > 1.116 mrg 361:/* FALLTHROUGH */ > Yes, I fixed it a few minutes ago. Martin
Re: CVS commit: [netbsd-9] src/sys
On Sun, 8 Mar 2020 10:54:43 + Martin Husemann wrote: > Pull up following revision(s) (requested by pgoyette in ticket #762): > cvs rdiff -u -r1.121 -r1.121.2.1 src/sys/arch/amd64/amd64/trap.c Hi, Someone reported on IRC that this seems to have become mismerged: /usr/src/sys/arch/amd64/amd64/trap.c: In function 'trap': /usr/src/sys/arch/amd64/amd64/trap.c:360:2: error: expected expression before '}' token 1.107 christos 356: 1.121.2.1! martin 357:MODULE_HOOK_CALL(amd64_oosyscall_hook, (p, frame), ! 358:ENOSYS, hook_ret); ! 359:if (hook_ret == 0) 1.107 christos 360:} 1.116 mrg 361:/* FALLTHROUGH */
