summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2015-01-13Initialize d_type to DTYPE_SMD in the default label to avoid disklabel(8)Miod Vallat
asking for the device type upon labeling a freshly formatted disk.
2015-01-13Generate a PT_PHDR entry for static PIE binaries. Modern GDB needs this toMark Kettenis
do some sanity checking while it determines the load base. ok kurt@
2015-01-13adapt mac.c to ssherr.h return codes (de-fatal) and simplify dependenciesMarkus Friedl
ok djm@
2015-01-13sync changes from libopenssh; prepared by markus@Damien Miller
mostly debug output tweaks, a couple of error return value changes and some other minor stuff
2015-01-13sync changes from libopenssh; prepared by markus@Damien Miller
mostly debug output tweaks, a couple of error return value changes and some other minor stuff
2015-01-13Many architectures call initmsgbuf() really really early, before uvm isMark Kettenis
initialized. Calling malloc(9) at that point is not a good idea. So initialize consbuf later. Fixes dmesg -s on sparc64 (and probably a few other architectures). ok miod@, deraadt@
2015-01-13zap really odd code that's not actually in-use.Marc Espie
okay millert@
2011-04-24Here comes the easter bunnytils 2.17 (the last version released against aMiod Vallat
licence mere mortals can understand the terms of); will be connected to the build on an arch-by-arch basis. Testsuites and generated files have been intentionnaly omitted from this import. Peer pressure and ok from at least drahn@ pirofti@ deraadt@
2015-01-13For non-blocking sockets tls_connect_fds() could fail with EAGAIN.Alexander Bluhm
Use the same logic from the read, write, accept functions to inform the caller wether a readable or writable socket is needed. After that event, the connect function must be called again. All the checks before connecting are done only once. OK tedu@
2015-01-13Remove a variable that's not really being used.Brian Callahan
ok deraadt@ jasper@
2015-01-13Always allow abort tasks to be scheduled, even if the device is beeingMartin Pieuchot
detached, in order to prevent a deadlock situation. This situation can occur if the thread detaching a device is sleeping, waiting for all submitted transfers to finish, and the device's pipes have not yet been aborted. This can happen when a USB Ethernet device is being detached while a userland program is doing an ioctl(2). Abort tasks need to be able to run in such case since timed out transfers rely on them to be properly completed. ok deraadt@
2015-01-13bump failure to set the initial time from debug to warningBrent Cook
from Paul B. Henson, ok phessler@
2015-01-13implement a SIGINFO handler so we can discern a stuck fuzz testDamien Miller
from a merely glacial one; prompted by and ok markus
2015-01-13Document the openssl s_client -proxy feature.Alexander Bluhm
OK jmc@
2015-01-13remove .Tn from stdio manuals; Kaspars Bankovskis found one of theseIngo Schwarze
2015-01-13do not spawn a pager when there is no output; issue pointed out by deraadt@Ingo Schwarze
2015-01-13Only notify userland about resolved ARP entries if the Ethernet addressMartin Pieuchot
changed or if we asked for it. Should reduce the RTM_RESOLVE storm seeing by claudio@ With inputs from and ok mikeb@
2015-01-13Restore sending RTM_ADD and RTM_DELETE messages to userland for everyMartin Pieuchot
route created/deleted with rt_ifa_add(9)/rt_ifa_del(9), not only for RTF_LOCAL routes. Regression introduced in r1.172 when restoring the original behavior of RTM_NEWADDR/RTM_RTM_DELADDR reported by Florian Riehm. Joint work with Florian Riehm, with input from and ok bluhm@.
2015-01-13Add the possibility to use the openssl s_client tool with an httpAlexander Bluhm
proxy. Implement the -proxy feature in the same hackish way as -starttls. OK jsing@
2015-01-13Add dmesg -s support, to view the output of rc(8) system startup messages.Marco Pfatschbacher
Help and feedback by Theo and Miod. OK deraadt@, manpage-ok jmc@
2015-01-13bump copyright yearReyk Floeter
2015-01-13bump copyright yearReyk Floeter
2015-01-13Abort if fcgi_chunked is not true to avoid sending additional garbageReyk Floeter
after the response. Found by Erik Lax ok florian@
2015-01-13syncTheo de Raadt
2015-01-13libedit claims compatibility with readline 4.2, so provide a stub forReyk Floeter
rl_set_keyboard_input_timeout() that was added in this version. Also crank the minor version. OK krw@ yasuoka@ deraadt@
2015-01-13use $SSH instead of installed ssh to allow override; spotted by markus@Damien Miller
2015-01-13regress test for PubkeyAcceptedKeyTypes; ok markus@Damien Miller
2015-01-13add sshd_config HostbasedAcceptedKeyTypes and PubkeyAcceptedKeyTypesDamien Miller
options to allow sshd to control what public key types will be accepted. Currently defaults to all. Feedback & ok markus@
2015-01-13Remove unnecessary calls to __atexit_register_cleanup(), calling __sinit()Philip Guenther
instead where necessary. Based on a diff from enh (at) google.com ok millert@
2015-01-13escape "No" when it's not intended as a macro, and zapJason McIntyre
trailing whitespace;
2015-01-13An article on medium.com highlighted that grepDaniel Dickman
was available in v4. To quote Ken Thompson, grep appeared "sometime before the 4th edition." ok schwarze@
2015-01-13-/* Backwards compatibility. */Theo de Raadt
-#ifndef NI_MAXHOST -#define NI_MAXHOST 1025 -#endif Disappointing.
2015-01-13"wi0" should be "an0" in the EXAMPLES section.Lawrence Teo
2015-01-13Give tcpdrop a proper usage() like other programs in the tree.Lawrence Teo
ok bluhm@
2015-01-13Implement more thorough error checks:Lawrence Teo
- Check the return value of every relevant function call. - If BIO_new() returns NULL instead of a valid BIO, do not attempt to blindly use the NULL value as a BIO throughout the rest of the code. - Ensure that bio_out is freed by BIO_free_all() at the end of all error paths. ok doug@
2015-01-13fix some memory leaks in dns handling.Brent Cook
- Nothing seems to free the result of host_dns(), so add host_dns_free() and call after each query. - If imsg_add() fails, it frees buf. Avoid subsequently dereferencing the freed buf in imsg_close(). ok millert@ deraadt@
2015-01-13pass the vnd xfer pointer to the tasks callback as part of theDavid Gwynne
vndbuf allocation. luke-warm support and ok krw@
2015-01-13prefer sizeof(thing) to sizeof(type)Brent Cook
ok tedu@ deraadt@
2015-01-13Enable snooping on Bay Trail HD Audio.Jonathan Gray
Fixes audio problems on an Asus J1800I-C reported by Jan Stary.
2015-01-13Add missing Utilite case in imx_platform_init_cons()Jonathan Gray
spotted by Patrick Wildt
2015-01-13for the install: target, use cmp as a rough attempt for avoiding repeatedTheo de Raadt
make install from Simon Nicolussi ok jsing tedu
2015-01-13syncTheo de Raadt
2015-01-12include NFSCLIENT. ok deraadtTed Unangst
2015-01-12syncTheo de Raadt
2015-01-12the BUG described at the bottom is not a bug. It is how it is supposed toTheo de Raadt
work. Move it up into the main body. ok tedu millert
2015-01-12lager -> larger (no beer involved)Miod Vallat
2015-01-12Fix two 19 years old (since day one) bug in bad144 sector reassignment logic:Miod Vallat
- when entering bad144 mode, correctly compute the actual C/H/S values for the logical block number. - when leaving bad144 mode and resuming normal I/O operation, when computing back the C/H/S values for the next logical block number, compute the sector number correctly.
2015-01-12Add fgetwln(3) from FreeBSD and bump libc minor revision.Todd C. Miller
2015-01-12Documentation for postgresql.port.mk.Vadim Zhukov
2015-01-12Make putting a breakpoint on the first instruction of a program work.Mark Kettenis
Previously a breakpoint at the first instruction would force us to single-step. Depending on the single-step implementation this would make us step over the breakpoint and continue in many cases. Fix this by passing the address of the current instruction pointer which will skip the code that checks if we need to step over a breakpoint. A similar fix has been made upstream many moons ago. Fixes debugging static PIE on sparc64 and probably others that use some sort of software single-stepping.