summaryrefslogtreecommitdiffstats
path: root/lib/libc (follow)
Commit message (Collapse)AuthorAgeFilesLines
...
* sigstack() is long dead, and the compat sigaltstack syscall is gone too.guenther2011-11-228-66/+15
| | | | | | Correct the namespace protections for sigreturn(), sigwait(), and psignal() ok millert@
* Complete the HISTORY of all syscalls that exist since Version 1 AT&T UNIX.schwarze2011-11-179-25/+118
| | | | | All facts from http://minnie.tuhs.org/cgi-bin/utree.pl, checked by sobrado@. Feedback and ok jmc@ jmc@ (sic, Jason checked and ok'ed this twice).
* Fix HISTORY:schwarze2011-11-171-6/+8
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Mention the predecessor exec() in v1, and remove the bogus "appeared in 3BSD", just leave "execve() appeared in v7", because: (1) There is a direct line of inheritance from v7 (Bell, Jan 1979) via 32v (Bell, May 1979) to 3BSD (UCB, Feb 1980), and it goes without saying that children include code from their grandparents. (2) It is impossible that there was parallel development of execve() at Bell and UCB. The only UCB release before v7 was 1BSD (May 1978), and even 2BSD (May 1979) did not include any kernel parts but fully relied on the Bell v6 kernel. When Bell released 32v, the UCB did not even own a VAX, and Bill Joy was still are pure userland hacker. Only after the original UCBVAX arrived at Berkeley, Richard Fateman involved Domenico Ferrari who involved Ozalp Babaoglu who started kernel work at UCB, based on Bell 32v, later involving Bill Joy. (3) Genetic analysis of the source code confirms this. All kernel code involved is in sys1.c. Regarding this file, the first two UCB releases, VAX 3BSD (based on Bell 32v) and PDP-11 2.8BSD (Dec 1981, based on Bell v7) are more similar to their respective Bell parents than to each other. The Berkeley versions are cousins, not siblings: v7 -> 32v: 523 +71 -60 = 534 v7 -> 2.8: 523 +305 -24 = 804 32v -> 3: 534 +169 -110 = 593 3 -> 2.8: 593 +413 -202 = 804 References: http://oreilly.com/catalog/opensources/book/kirkmck.html http://minnie.tuhs.org/cgi-bin/utree.pl?file=32V http://minnie.tuhs.org/cgi-bin/utree.pl?file=V7/usr/sys/sys/sys1.c http://minnie.tuhs.org/cgi-bin/utree.pl?file=32V/usr/src/sys/sys/sys1.c http://minnie.tuhs.org/cgi-bin/utree.pl?file=3BSD/usr/src/sys/sys/sys1.c https://www.mckusick.com/csrg/ - /cdrom1/2.8/usr/kernel/sys/sys/sys1.c facts checked and ok sobrado@, style and formatting ok jmc@
* Fix mbstowcs return value documentation. New text based on mbsrtowcs page.stsp2011-11-161-16/+19
| | | | ok jmc
* Make __svfscanf() the unlocked, core of vfscanf() and use it inguenther2011-11-086-46/+23
| | | | | | sscanf()/vsscanf() where locking is unnecessary. ok millert@
* Expand HISTORY; feedback and ok jmc@ sobrado@.schwarze2011-11-071-8/+21
| | | | | | | | | | This one is tricky, so i'm giving the sources: http://minnie.tuhs.org/cgi-bin/utree.pl?file=V1/u2.s http://minnie.tuhs.org/cgi-bin/utree.pl?file=V4/nsys/ken/sys3.c http://minnie.tuhs.org/cgi-bin/utree.pl?file=V5/usr/sys/ken/sys4.c http://minnie.tuhs.org/cgi-bin/utree.pl?file=V6/usr/sys/ken/sys4.c http://minnie.tuhs.org/cgi-bin/utree.pl?file=V7/usr/sys/sys/sys4.c http://www.openbsd.org/plus21.html (lchown)
* More .At v1 HISTORY.schwarze2011-11-073-14/+39
| | | | | | All information from http://minnie.tuhs.org/cgi-bin/utree.pl . All facts checked by sobrado@. Feedback and ok on an earlier version jmc@.
* don't handle out-of-mem conditions using compiled out asserts (ugh). Fromotto2011-11-071-9/+10
| | | | netbsd; ok deraadt@
* Commemorate research!dmr by documenting the history of allschwarze2011-11-067-40/+53
| | | | | | | | | system calls existing since Version 1 AT&T UNIX (first batch). All information taken from primary sources at the UNIX tree of the UNIX heritage society, http://minnie.tuhs.org/cgi-bin/utree.pl . The number of errors in the existing manuals is astonishing. All facts checked by sobrado@. Feedback and ok on wording and formatting by jmc@.
* .Ev -> .Dvderaadt2011-11-064-16/+16
|
* repair documentation offset in the p*() versions of the calls and EINVAL;deraadt2011-11-062-27/+25
| | | | | plus merge differences between two pages. ok guenther jmc
* Move around the error cases to make clear that some don't apply to theguenther2011-11-052-27/+45
| | | | p* versions, as well as fix a couple other cases
* Sync scanf(3) to wscanf(3), and a few bits the opposite way:schwarze2011-11-022-66/+50
| | | | | .Fd -> .In, .Li -> .Vt, and remove various other gratuitous differences. feedback and ok stsp@, ok jmc@
* Allow segments to be used even after they were marked for deletion withrobert2011-10-271-4/+11
| | | | | | | | | | | | the IPC_RMID flag. This is permitted as an extension beyond the standards and this is similar to what other operating systems like linux do. Because compat_linux(8) was emulating this already, remove that code since now this is the default. input from oga@, guenther@, jmc@, deraadt@ ok deraadt@
* Update POSIX references for chdir(2) and chmod(2) from 1988 to 2008.schwarze2011-10-244-15/+29
| | | | | | | | Add POSIX references for fchdir(2), fchmod(2), fchmodat(2), fchown(2), lchown(2) and execve(2). State that fchmod(2) can return EPERM as required by POSIX 2008 (tested). Requested by sobrado@; ok guenther@ jmc@ sobrado@.
* Bump libc major; wscanf and wcsftime were addedstsp2011-10-161-2/+2
|
* Hook wcsftime(3) to the build; committing on behalf of espiestsp2011-10-161-3/+3
|
* Add wscanf(3) and friends. Based on our scanf(3) implementation, with widestsp2011-10-1610-10/+1531
| | | | | character support changes based on code from FreeBSD. ok espie guenther; man page help from schwarze
* Tweak wording, to clarify that setrtable affects the routing tableguenther2011-10-161-6/+10
| | | | | | | and that getrtable returns the table, not 0/-1. Also, strip some quotes that mandoc made superfluous, per comment from jmc@. ok claudio@
* Note a restriction that is actually enforced in 5.0guenther2011-10-161-2/+3
|
* Fixes and tweaks from jmc@guenther2011-10-161-7/+7
|
* Make consistent the syscall stubs for the syscalls that got specialguenther2011-10-1610-23/+83
| | | | | | | | handling to fix up the alignment of 64bit arguments so that they do the same dance where _thread_sys_FOO is the real stub and FOO is a weak alias. For some of them, this is needed for cancellation handling. From discussions with fgsch@, ok millert@
* Add stubs and manpage for __{get,set}_tcbguenther2011-10-162-3/+68
| | | | ok deraadt@
* Fix toprec handling in cgetnext(). OK otto@millert2011-10-141-4/+9
|
* Change to syscall is done, so delete an XXX commentguenther2011-10-111-5/+1
|
* put in code for wcstfime, as discussed with millert@ and stsp@ (and testedespie2011-10-102-0/+605
| | | | by ajacoutot@, thx), to be linked in and activated alongside wscanf...
* Stop defining XPG4_1994_04_09.ajacoutot2011-10-101-2/+1
| | | | | | | | | | | | | | | | As kettenis@ pointed at, http://pubs.opengroup.org/onlinepubs/009695399/functions/strftime.html has the following: The Open Group Corrigendum U033/8 is applied. The %V conversion specifier is changed from "Otherwise, it is week 53 of the previous year, and the next week is week 1" to "Otherwise, it is the last week of the previous year, and the next week is week 1". In effect, strftime(3) with the "%V" format will now return "52" instead of "53" for days in January before the first Monday, when January 1st falls on a Friday, Saturday, or Sunday. ok kettenis@ millert@
* d_t_fmt: fix the format for the day of the month to use %e (singleajacoutot2011-10-091-2/+2
| | | | | | digits are preceded by a blank instead of a 0) and not %d (POSIX). ok jasper@ guenther@
* Make struct chunk_info a variable sized struct, wasting lessotto2011-10-061-28/+55
| | | | | space for meta data by only allocating space actually needed for the bitmap (modulo alignment requirements). ok deraadt@
* signedness fix; ok millertdtucker2011-10-021-2/+2
|
* KERN_BUFCACHEPERCENT defaults to 20%, not 10%; from Juha Erkkilajmc2011-09-251-3/+3
| | | | ok ariane
* Stop using the latin1 character set in the "C" locale. This behaviour wasstsp2011-09-221-18/+17
| | | | | | | non-standard. Tools which aren't locale-aware should only print ASCII but were also printing latin1 which looks garbled in UTF-8 terminals. Folks who want to use an extended character set now must set LC_CTYPE. ok millert uwe naddy espie
* fix GLOB_KEEPSTAT without GLOB_NOSORT; the implicit sort was beingdjm2011-09-221-6/+44
| | | | | applied only to the gl_pathv vector and not the corresponding gl_statv array. reported in OpenSSH bz#1935; feedback and okay matthew@
* In glob(3), limit recursion during matching attempts. Similar to fnmatch fix.stsp2011-09-201-5/+17
| | | | | Also collapse consecutive '*' (from NetBSD). ok miod deraadt
* Make this actually work.kettenis2011-09-191-4/+4
| | | | ok jsing@
* Use a long-displacement branch instruction to call __cerror, since thekettenis2011-09-191-2/+3
| | | | linker doesn't generate stubs yet.
* Clear the upper 32-bits of the system call address we load using ldil, sincekettenis2011-09-191-2/+4
| | | | | that instruction does sign extension. Look at %r1 to determine whether a system call was sucessful, since that is where the kernel puts the information.
* arithemtic -> arithmeticmiod2011-09-171-2/+2
|
* after some thought, i think it unlikely that we will need to specifyjmc2011-09-162-8/+8
| | | | | | | | | | a revision number for ISO 8601; in addition, the revision number makes the text unwieldy when discussing the standard in a casual manner. therefore change the macro from -8601-4 to simply -8601, which will output "ISO 8601". i'll get kristaps or ingo to make the change upstream to mandoc. discussed with/ok millert
* use .St here for ISO 8601 references; ok millertjmc2011-09-162-9/+21
|
* RFC-2822 -> RFC 2822jmc2011-09-161-3/+3
|
* update vt100 example to vt200, as vt100 apparently is no longer in thejmc2011-09-141-6/+6
| | | | | | | termcap database; from Nils Anspach ok nicm
* Preserve errno across calls to open the password database(s), soguenther2011-09-051-2/+5
| | | | | | | that errno isn't changed when a normal user (who can't open spwd.db) does a lookup. Problem pointed out by Tim van der Molen (tbvdm at xs4all.nl)
* Switch sigreturn() to the normal syscall entry instead of int$80.guenther2011-09-052-26/+3
| | | | | | This was done for the sigreturn call in sigcode before 5.0. ok deraadt@ for post-5.0
* knock out some useless Pp;jmc2011-09-032-4/+2
|
* some minor tweaks;jmc2011-09-032-13/+8
|
* make -column lists pretty again;jmc2011-09-0311-653/+716
| | | | | | | | | | specifically, rewrite them to permit some markup in the column headers, and use "Ta" instead of literal tabs; mandoc does not currently match groff 100%, but a mandoc fix may be some time off, and we've gone enough releases with poorly formatting column lists. in some cases i have rewritten the lists as -tag, where -column made little sense.
* only one leap second is permitted, not two; from guentherjmc2011-08-312-8/+8
|
* in column lists, enforce the following ordering:jmc2011-08-302-6/+6
| | | | | | | | | .Bl -column header1 header2 ... <Bl flags> mixing things up confuses mandoc, and adding flags in can screw up the list. this hopefully just makes things a bit safer (and more consistent). (one more bit of work left and all column lists should format how we want)
* as with other list types, column lists generally do not need a Pp/-compactjmc2011-08-191-4/+3
| | | | | | | | | | | | | | | | | construct; this also sidesteps what seems to be a problem with mandoc, in that "-column -compact" seems to mess up the formatting. thus these pages should now have their lists formatted nicely (i.e. correctly aligned and with indent applied); as a side note, the fact that headers are not properly marked up is another issue which will be addressed separately (a mandoc fix is needed, i think). i have fudged a few of these to mark up properly, since the workaround does make sense for some pages. as another side note, i haven;t fixed man7, as i need to prepare a separate diff for kristaps and ingo.
Copyright © 1996 – 2023 Jason A. Donenfeld. All Rights Reverse Engineered.