Date: Sat, 28 Mar 2020 11:46:29 +1100 From: matthew green <m...@eterna.com.au> Message-ID: <15233.1585356...@splode.eterna.com.au>
| can we just leave this as-is and let netbsd GCC people care? Only if the GCC people do care, and understand the issue, and implement what we want.... Based entirely upon messages here, it seems as if they believe exactly the opposite (or once did), eg: one of my systems has ... -rw------- 1 root wheel 0 Aug 1 2019 /var/tmp/cc0CF9Uh.le -rw------- 1 root wheel 0 Jun 15 2019 /var/tmp/cc0SLuqk.c -rw------- 1 root wheel 94702 Aug 1 2019 /var/tmp/cc0uAn0K.ld -rw------- 1 root wheel 0 Jun 1 2019 /var/tmp/cc2wYgEG.o -rw------- 1 root wheel 0 May 21 2019 /var/tmp/cc2zHkbz.le -rw------- 1 root wheel 0 Apr 29 2019 /var/tmp/cc4G3m88.le -rw------- 1 root wheel 98595 May 21 2019 /var/tmp/cc4dzxVs.ld -rw------- 1 root wheel 98595 Jun 3 2019 /var/tmp/cc5MqdQs.ld -rw------- 1 root wheel 98595 Jun 1 2019 /var/tmp/cc5ROcTW.ld -rw------- 1 root wheel 0 Aug 1 2019 /var/tmp/cc5sAGZG.c -rw------- 1 root wheel 94702 Jul 10 2019 /var/tmp/cc73am4Y.ld -rw------- 1 root wheel 0 Apr 22 2019 /var/tmp/ccAJcqpq.le -rw------- 1 root wheel 0 Aug 1 2019 /var/tmp/ccCHeiPd.o -rw------- 1 root wheel 0 Jul 10 2019 /var/tmp/ccDZoY7q.c -rw------- 1 root wheel 0 Apr 29 2019 /var/tmp/ccEDt7pv.le -rw------- 1 root wheel 0 Aug 1 2019 /var/tmp/ccGG3Duh.le -rw------- 1 root wheel 0 Apr 18 2019 /var/tmp/ccIAAB5X.o -rw------- 1 root wheel 0 Jul 10 2019 /var/tmp/ccIB2j2f.le -rw------- 1 root wheel 0 Apr 18 2019 /var/tmp/ccJOlV73.c -rw------- 1 root wheel 0 Jun 3 2019 /var/tmp/ccJgigog.c -rw------- 1 root wheel 0 Jun 15 2019 /var/tmp/ccKW5VdP.le -rw------- 1 root wheel 0 May 1 2019 /var/tmp/ccMI998l.o -rw------- 1 root wheel 0 Apr 29 2019 /var/tmp/ccOLPHE9.o -rw------- 1 root wheel 98595 Apr 29 2019 /var/tmp/ccP3twoE.ld -rw------- 1 root wheel 0 Jun 3 2019 /var/tmp/ccSJQiCm.o -rw------- 1 root wheel 0 Jun 3 2019 /var/tmp/ccUfI43y.le -rw------- 1 root wheel 0 May 21 2019 /var/tmp/ccWJa4pg.c -rw------- 1 root wheel 0 Apr 29 2019 /var/tmp/ccYvThNk.o -rw------- 1 root wheel 0 Aug 1 2019 /var/tmp/cca0jZaH.c -rw------- 1 root wheel 0 Apr 22 2019 /var/tmp/cccefbii.o -rw------- 1 root wheel 0 May 1 2019 /var/tmp/cceQHHTx.le -rw------- 1 root wheel 0 Jun 1 2019 /var/tmp/ccf4Nepq.c -rw------- 1 root wheel 98595 Apr 29 2019 /var/tmp/ccg0gH6p.ld -rw------- 1 root wheel 0 Apr 29 2019 /var/tmp/cch9cpUE.c -rw------- 1 root wheel 0 May 21 2019 /var/tmp/ccia5LFm.o -rw------- 1 root wheel 98595 May 1 2019 /var/tmp/cclxyp1r.ld -rw------- 1 root wheel 98595 Apr 18 2019 /var/tmp/ccnS4e3R.ld -rw------- 1 root wheel 0 Apr 29 2019 /var/tmp/cco6yYtf.c -rw------- 1 root wheel 0 Aug 1 2019 /var/tmp/ccoXNE5d.o -rw------- 1 root wheel 0 May 1 2019 /var/tmp/ccpQBE9f.c -rw------- 1 root wheel 0 Jun 15 2019 /var/tmp/ccsxBUGu.o -rw------- 1 root wheel 0 Apr 22 2019 /var/tmp/cctaeuee.c -rw------- 1 root wheel 0 Apr 18 2019 /var/tmp/ccukTN0L.le -rw------- 1 root wheel 98595 Apr 22 2019 /var/tmp/ccvEGOlm.ld -rw------- 1 root wheel 0 Jun 1 2019 /var/tmp/ccwhr37c.le -rw------- 1 root wheel 93664 Aug 1 2019 /var/tmp/ccxwbXEK.ld -rw------- 1 root wheel 98595 Jun 15 2019 /var/tmp/ccyaWpXE.ld -rw------- 1 root wheel 0 Jul 10 2019 /var/tmp/ccyxx75H.o Those will be from system builds most likely interrupted by power fails (no newer ones as I stopped doing builds there - that there are no older ones suggests that the last time I went and manually cleaned up was March or early April last year). Trash like this that either needs manual cleaning, or purpose added scripts, for no good reason, is absurd. Even the (no longer true anyway, but once was) "more space available in /var/tmp" is (and always was) stupid for this purpose, even on the most ancient of systems, /tmp was always made big enough for compiler temp files, which aren't really ever all that huge - and while today we might have lots of parallel compiles running, we also have bigger /tmp, the ancient systems with 10Mb /tmp partitions couldn't really cope with more than one or two compiles happening simultaneously (and certainly wouldn't with the CPU/mem load imposed by a modern gcc). So, if the gcc people are willing to (or maybe already have) switched to /tmp for temporary files (in a way that can be overridden by setting TMPDIR) then of course, we should just use their version. If not, we need to fix it. And: m...@eterna.com.au said (in a different message): | i don't like this "don't care about netbsd-8" feeling i'm hearing I think you took Martin's words out of context. I read that as applying only to where gcc puts its temp files in NetBSD-8 (which is the wrong place, on NetBSD anyway) - but which, after several years of doing that, is not exactly crucial to fix (and is the kind of churn that isn't wanted or needed for supposedly stable systems). kre