summaryrefslogtreecommitdiff
path: root/usr.bin/make
AgeCommit message (Collapse)Author
2013-07-07old option -P bites the dust. It's not been doing anything for years.Marc Espie
It used to control pipes for parallel output, but the new model means pipes are no longer used at all. Its mere presence confuses some people. tested thru a few builds and bulk to not impact anything.
2013-06-24do not use literal displays when the display contains mark up;Jason McIntyre
2013-06-20fix typoMarc Espie
thx Caspar Schutijser
2013-06-15make explicit some hidden assumptions, make path handling clearer,Marc Espie
fix set -e description, as we actually match what posix says. a bit of feedback from millert@, guenther@, jmc@ will probably have further fixes when he's back :)
2013-05-30remove cmtime again, but with a proper test for nodes without children.Marc Espie
problem seen by aja, make pointed by matthieu, sleuthing by me, okay by millert (and you say OpenBSD developers don't work together)
2013-05-25obviously missed something, so revert.Marc Espie
(noticed by sthen@/matthieu@)
2013-05-22as checked through thorough tests, youngest->mtime == ctime, so ditchMarc Espie
the extra field. remove some extra abstraction layer: use clock_gettime directly instead of ts_set_from_now (what is "now" anyways) time_to_string takes param by pointer rename "now" into starttime (more accurate term) randomize queue uses arc4random_uniform (prompted by deraadt@) display debug timestamp with ns too (it's debug, so it doesn't really matter whichever way it's done, as long as it's done) okay millert@
2013-05-14keep track of the youngest child, helps a lot with out-of-date messagesMarc Espie
in -dm mode. okay millert@
2013-05-06clarify that the :C modifier uses extended regular expressions; ok espie@ jmc@Christian Weisgerber
2013-05-06actually show list of mismatched for loops when a fatal error occurs.Marc Espie
okay millert@
2013-04-23remove TIMESTAMP abstraction layer, prodded by theo.Marc Espie
while there, clean up includes. use strtoll for ar timestamps (pretty much unused in reality, more standard conforming than anything) use idea from Todd to adapt to time_t being 32 bits OR 64 bits (pedantically correct: INT_MIN would work just fine up to 1910 or so...) okay millert@, gone thru a make build.
2013-04-22unifdef non full TIMESPEC codeMarc Espie
2013-02-19Add missing trailing slash on the :S modifier in the BUGS section.William Yodlowsky
ok espie
2013-02-15turns out if a line ends with a $, the var module will run off the cliffMarc Espie
while expanding it. shows up very prominently with MALLOC_OPTIONS=S Problem found by jasper@ okay'd by jasper@ and kind-of duh'd by miod@
2013-02-04synch with reality, okay jmc@Marc Espie
2012-12-22cosmetic changes:Marc Espie
- pref -> prefix as pref is somewhat ambiguous - rework tests so loop doesn't intend off the screen - zap unneeded variable okay krw@
2012-12-14a few more commentsMarc Espie
2012-12-08document a bit of job.cMarc Espie
notice that Job_Finish() really returns a boolean, so unconfuse that accordingly (it's likely the extra Fatal() message is not needed and we could just call finish)
2012-12-07job.h is also used for compat sequential mode.Marc Espie
2012-12-07document engine interface, remove internal function from visible interfaceMarc Espie
2012-12-07gc dead codeMarc Espie
2012-12-07gc old defineMarc Espie
2012-12-06missing externMarc Espie
2012-12-06... and zap redundant declaration.Marc Espie
2012-12-06old code is gone, remove comments and prototypesMarc Espie
2012-11-24complement "need an operator" fatal error message with the actual line content.Marc Espie
okay beck@
2012-11-24Don't output a . at the end of directory, as it is confusing.Marc Espie
okay beck@
2012-11-24be more strict in recognizing .if keyword() constructs, don't ignoreMarc Espie
unknown keywords. uniform white space handling. okay beck@
2012-11-24clean-up: move up .USE handling, so that the rest becomes a simple switch()Marc Espie
okay beck@
2012-11-21${.ALLSRC} and ${.OODATE} should always be defined, even for empty listsMarc Espie
of prerequisites. This prevents complaints from the var module and from other developers. That's a bug I introduced 5 years ago... found out by miod@ okay miod@
2012-11-07say goodbye to killing trailing spaces.Marc Espie
this was yet another weirdness in our make that isn't shared by other makes, and that isn't part of any standard. This means end-of-line spaces in variables ARE significant (spaces around the equal sign still aren't). okay sthen@, "sounds fine" deraadt@
2012-10-23fix fringe case where length is around 140.Marc Espie
problem found by Tobias Ulmer.
2012-10-18numerous error message fixes:Marc Espie
- do ^C checking differently: don't record sent signals, but when jobs die, recheck whether we received/have pending a INT/QUIT/TERM/HUP signal. Then don't display our process group "normally", instead group together everything dying by signal/shell dying by signal (just give the target names). - make certain we always handle signals before dying from "other conditions" - have the parser messages look more like normal messages - remove double error messages from some parser errors - make sure unclosed variables ARE errors when some modifiers are present - keep track of the base directory we're run from, so that submakes can get shortened directories... - make sure the whole error message including silent command fits into a reasonable length. okay millert@
2012-10-12Document duplicate inference rules. That's posix behavior, and what we do.Marc Espie
2012-10-12fix infinite loops on regexps that match the empty word, which isMarc Espie
especially bad during runtime as we trap ^C. For instance ${A:C/-*//g}. Problem found by rpe@ Do the same as other modern tools: advance by one char and repeat, so that there's at most one empty word match at each position. okay millert@
2012-10-11amend bug section with some fixed stuff.Marc Espie
2012-10-11fix make's cond parser to be able to handle .if 5 < 7 directly.Marc Espie
okay millert@
2012-10-09... yet more documentation, lots of inputs from jmc@ as usual.Marc Espie
2012-10-09steal .if commands() concept from NetBSD.Marc Espie
Actually less ambiguous than .if target().
2012-10-09warn about targets with multiple command lists (debug option), as it'sMarc Espie
definitely non portable behavior. discussed with millert@, who rightfully insisted on the optional debug part
2012-10-09actual engine change: prevent jobs from stomping on each other's files.Marc Espie
namely: - targets that ARE the same file shouldn't build concurrently. - targets in the same group (e.g., file.c file.c: file.y) shouldn't build concurrently. This probably fixes all the remaining races in make -j4 build in src.
2012-10-09recognize new debug options for double commands, heldjobs, target groupMarc Espie
construction
2012-10-09tweak error messages yet some more, lots of feedback fromMarc Espie
deraadt@, millert@
2012-10-09- SPECIAL_DEPRECATED -> SPECIAL_NOTHINGMarc Espie
- nodes for .POSIX and .SCCS_GET (which don't do anything) - zap remaining suffix crud. new scaffolding: - groupling list and HELDBACK state to avoid races in engine. - parser recognizes lists of targets that shoul be grouped together - OP_DOUBLE to mark nodes that have multiple lists of commands
2012-10-09new interface: look in command lines for variables that indicate a list ofMarc Espie
targets is actually separate targets.
2012-10-06truncate silent command output to two lines. okay millert@Marc Espie
2012-10-06rework section about special targets/prereq to be somewhat readable.Marc Espie
feedback & ok jmc@
2012-10-06- extra juice for debugging signal passing. Note when we can't pass theMarc Espie
signal because the process already bought it (pgroups will do that to you) (lots of discussion with Todd on that one) - tweak error handling some more to make it less verbose when just one job is running... - show signal name in case of signal interrupts. - zap OP_LIB, move that stuff to the location where we warn when we meet that bug. okay millert@
2012-10-05document implicit rules and target dependency operators better.Marc Espie
some input from jmc@, might not be 100% perfect yet, but better than nothing at all.
2012-10-04backout pgroup/job control from make, there is something deeply bogusMarc Espie
in stdin interaction. Fixes update-patches as reported by aja...
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-24 09:43:04 +0000