summaryrefslogtreecommitdiffstats
path: root/lib
Commit message (Collapse)AuthorAgeFilesLines
* Oops, fix logic for previous commitache1997-08-141-5/+5
|
* Remove wrong machine-dependent phrase about maximum microsecondsache1997-08-141-6/+5
| | | | limit. Small cleanup.
* Break loop if error != EINTR occurse or it can take forever withache1997-08-141-2/+5
| | | | time_remaining unchanged
* Define NCURSES_VERSION publicly, some applications want itache1997-08-135-37/+80
| | | | | Fix unctrl() Merging from 4.1 used for this fixes.
* Protect the copyright comments from reformatting bysteve1997-08-135-3/+28
| | | | | | | indent and make this compile -Wall clean like the Makefile suggests that it should. :) Pointed out by: Bruce Evans <bde@zeta.org.au>
* Note that default behaviour (restarting signals) is only for signal(3)ache1997-08-131-1/+4
|
* Handle syscalls error return slightly betterache1997-08-131-6/+12
|
* Back out itimerfix hack since nanosleep1 fixed nowache1997-08-131-58/+19
| | | | Handle syscalls error return slightly better
* Add unsigned char cast to all ctype callsache1997-08-131-27/+27
|
* Fix longstanding bug with buffer pointer goes beyoud buffer startache1997-08-131-1/+1
| | | | Cause initscr (ncurses) fail in some cases
* TRACE_VIRTPUT (we don't have it yet) -> TRACE_CHARPUTache1997-08-131-1/+1
|
* Rearrange itimerfix loop doing it inside SIGALRM handling sectionache1997-08-121-19/+20
|
* Solve itimerfix() problem completely by using loop in 100000000 secs chunksache1997-08-121-0/+19
|
* Reflect blocked SIGALRM changesache1997-08-122-3/+3
|
* Back out my installation of SIGALRM handler even if it is blockedache1997-08-122-38/+58
| | | | | | | and return to previous Peter's variant. POSIX says that this place is implementation defined and old variant allows application block SIGALRM and sleep and not be killed by external SIGALRMs. BTW, GNU sleep f.e. sleeps forever in blocked SIGALRM :-)
* Run mkMakefile.sh to regeneratepeter1997-08-122-124/+3
|
* Reflect -current implementationsache1997-08-122-20/+24
|
* 1) Make usleep() terminateable by alarm (or ualarm) as supposedache1997-08-121-4/+8
| | | | | historically 2) Fix end of time loop condition: && -> ||
* Change to reflect -current sleep implementation realityache1997-08-121-28/+13
|
* Make itimerfix hack better: return the time we can't sleep at onceache1997-08-121-12/+20
|
* Fix file descriptor leak.brian1997-08-114-2/+11
| | | | | Submitted by: Charles Mott <cmott@srv.net> Identified by: Gordon Burditt
* Default size for FD_SETSIZE is 1024 not 256.steve1997-08-111-1/+1
| | | | | PR: bin/4177 Submitted by: Wolfgang Helbig <helbig@MX-BA.Stuttgart.De>
* Implement canonical locking protocolache1997-08-103-62/+83
| | | | Suggested by: joerg
* Hack^H^H^H^Hworkaround for itimerfix(9) gratuitously limiting thejoerg1997-08-101-0/+16
| | | | | | | | | acceptable range for tv_sec to the magic number 100000000 (which at least ought to be declared in a header file, and explained in the non-existing man page, as well as in the existing man pages for nanosleep(2) & Co.). PR: bin/4259
* Import strptime(3) into libc. We've got permission by Kevin Ruddy tojoerg1997-08-097-495/+646
| | | | | | | | modify the original `no modifications' copyright message, and i've included his mail into the source file. The common localization functions between strptime(3) and strftime(3) have been broken out into timelocal.[ch].
* This commit was generated by cvs2svn to compensate for changes in r28019,joerg1997-08-091-0/+358
|\ | | | | | | which included commits to RCS files with non-trunk default branches.
| * Clean import of strptime(3) onto a vendor branch.joerg1997-08-091-0/+358
| | | | Submitted by: Powerdog Industries <kevin.ruddy@powerdog.com>
* Observe precedence set by Phillippe Charnier in adding ansteve1997-08-074-12/+20
| | | | rcsid.
* Remove #if(n)def BSD_4_4_LITE cruft and sccsid -> rcsid.steve1997-08-074-8/+12
|
* Improve weak locking by using flock()ache1997-08-051-4/+8
|
* Update to version 2.2. Only the PacketAlias*()brian1997-08-039-196/+1071
| | | | | | functions should now be used. The old 2.1 stuff is there for backwards compatability. Submitted by: Charles Mott <cmott@snake.srv.net>
* Remove collate_range_cmp, was left for temp. backward compatibilityache1997-08-031-10/+1
|
* Implement dlsym(RTLD_NEXT, symbol).jdp1997-08-023-8/+42
|
* Move tcl packages to libdata/pkgphk1997-08-012-3/+3
|
* Fix handling of mixed colors+attributes case by merging from ncurses 4.1ache1997-07-301-7/+21
|
* Fix logical background handling by merging it from ncurses 4.1ache1997-07-3014-37/+102
| | | | No new user-visible functions added
* ifdef out ttytype definition, mytinfo not have it and configure confusedache1997-07-301-0/+2
|
* Improve dependancy behaviour a little more.msmith1997-07-281-3/+4
| | | | | | This is still pretty disgusting. Submitted by: bde
* Add getbkgd() macroache1997-07-271-0/+1
|
* Remove a reference to /c/phk, this should be handled by the magic scriptphk1997-07-271-1/+1
| | | | in src/tools.
* List help.h a a dependancy for editline.cmsmith1997-07-271-2/+2
| | | | | | | | Use 'beforedepend' instead of '.depend' to hang automatically-generated headers off. XXX the latter is bogus without a 'beforeall' target and explicit ordering of dependancy generation for targets.
* Fix a minor typo. It only affects the sparc version.jdp1997-07-261-2/+2
|
* Added bounds checking to the example after chasing down someone'spds1997-07-261-1/+2
| | | | | broken code where they'd copied the example basically verbatim and blew an array. (engage brain before typing 'make')
* Tada!, tcl is now at 8.0b2 I belive.phk1997-07-252-68/+166
|
* Reorder decomposition of the filename argument for the '-h' flag tomsmith1997-07-231-2/+3
| | | | | | | avoid misprocessing in the case where the filename argument contains more than one period. Submitted by: micke@dynas.se (Mikael Hybsch)
* = -> ==, strcpy -> strncpy from OpenBSD.charnier1997-07-222-8/+8
| | | | | update man page. Add usage(). Obtained from: OpenBSD
* Add to CLEANFILES instead of setting it absolutely. Cleaning of *.S andbde1997-07-211-1/+1
| | | | tags was broken.
* sleep() after sending 'nologin' file to ensure output is drained beforedavidn1997-07-191-1/+2
| | | | disconnect.
* Make sleep() and usleep() "eat" any stray SIGALRM signals during thepeter1997-07-182-49/+29
| | | | | | | | lifetime of the call, just like the old implementation did. Previously, we were only eating them if the application did not call sleep()/usleep() with SIGALRM masked. Submitted by: ache
* Add appropriate ${DESTDIR} in front of absolute paths.asami1997-07-183-3/+3
|
OpenPOWER on IntegriCloud