summaryrefslogtreecommitdiff
path: root/libexec
AgeCommit message (Collapse)Author
2001-09-25Zero out dl_data before populating it.Artur Grabowski
Cleaner populating of dl_data.
2001-09-25Do mmap the right way.Artur Grabowski
2001-09-25Another attempt at getting this right. This time, play safe.Artur Grabowski
2001-09-24Some cleanup.Artur Grabowski
2001-09-24Some voodoo for bootstrapping sparc64.Artur Grabowski
2001-09-24Get the bootstrapping right.Artur Grabowski
2001-09-24Proper RELOC_RELA for bootstrap.Artur Grabowski
2001-09-24sparc64 needs slightly different flags.Artur Grabowski
2001-09-24Adapt to _dl_boot_bind change.Artur Grabowski
2001-09-24API change between asm code and _dl_boot_bind(). Simply andDale Rahn
reduce number of parameters. Alpha changes coming shortly.
2001-09-24dl_data is at the start of the stack, not after env. load loff.Artur Grabowski
Now _dl_boot starts correctly.
2001-09-23pass appropriate compilation flags for sparc64. needs to be compiled PIC.Dale Rahn
both C and asm files.
2001-09-23add munmap syscallDale Rahn
change _rtld references to _dl to match C code. change relocation types in archdep.h to be sparc64 relocs. rtld_machine.c copied from alpha, with some modes to compile for sparc64 This is in-tree development.
2001-09-22Upon futher contemplation, the notion of unmaphints is just wrong.Dale Rahn
First the maphints code is not cleaning up after itself, by leaving a file descriptor open and trying to clean up after the dynamic linker is done loading by removing data needed during dlopen() is nonsense. the _dl_unmaphints() function and it's references have been removed. _dl_maphints() now cleans up it's file descriptor before returning. static data used later during dlopen() is left intact.
2001-09-22Do not check for ':' twice, otherwise the rpath loses the first characterDale Rahn
after the : in the list of paths.
2001-09-22The unmaphints routine should not destroy the hint mapping.Dale Rahn
Otherwise shared modules which depend on libraries cannot find them because the list of trusted directories has been lost. However we need to be careful that the unmapping of the hints frees the file descriptor, but will not free it multiple times.
2001-09-21 Drop privs to named group, not group with same name as user.Bob Beck
noticed by itojun, thanks. fries, you need to catch this for what you're using with faithd
2001-09-21partially done... committing so other folks don't have to repeat some of the ↵Jason Wright
work
2001-09-21Stolen from alpha.Jason Wright
2001-09-21first whack at ldasm.S... this will not work yet (partially based on NetBSD,Jason Wright
partially based on alpha)
2001-09-19some signal race whackingTheo de Raadt
2001-09-19occured->occurredMike Pechkin
idea from deraadt@ via NetBSD millert@ ok p.s. Next commit will fix a typo in the sys/
2001-09-15Some cleanup in loader.c, initialize the symbol pointer with NULL,Dale Rahn
add a missing initialization of the sym pointer. Add some functionality which allows a program to open itself dlopen(NULL), so that it can then look up symbols in the executable itself. Note that the program can only access exported variables, either by exporting all variables with the ld option -E or externally referrenced. Fix bug in dlsym() where it would return failure when looking up symbols. It was testing the offset of the found symbol, not if the symbol was found.
2001-09-15IPv6 support from Ryan McBride (mcbride@countersiege.com)Mike Frantzen
2001-09-11krb5 before krb4 to please the linkerJason Wright
2001-09-06Initial idea from aaron@: Last char of .Xr group in SEE ALSO section shouldMike Pechkin
be a single digit. Powered by mantoya@. millert@ ok.
2001-09-05make sure that va_start() has matching va_end()Theo de Raadt
2001-09-05-Make ftp-proxy drop privs to user "proxy" by default.Bob Beck
*WARNING* this means that it will die when it can't find user proxy if you are not running with a passwd database generated from current - Speling cleanup and missing va_end() noticed by <d.doroshenko@omnitel.net> - fix logging of getpwnam|getgrnam failures.
2001-09-05Add SIGPIPE to the list of signals we catch and unlock/exit on.Todd C. Miller
2001-09-05s/pf_natlook/pfioc_natlook (ioctl parameter struct)Daniel Hartmeier
2001-09-04Use merr() not err() since due to function name change in mail.local/locking.cTodd C. Miller
2001-08-28remove unneeded code (this is done later)Bob Beck
2001-08-28oh, some cleanups etc etcTheo de Raadt
2001-08-27header orderTheo de Raadt
2001-08-23recieve -> receiveAaron Campbell
2001-08-22missing spaceBob Beck
2001-08-22oopsBob Beck
2001-08-22See also ftp-proxyBob Beck
2001-08-22Fixes for my worst abuses of nroff from mpech@Bob Beck
Add a sentence at the start for espie in dumb user mode :)
2001-08-22Duh.Bob Beck
2001-08-22-Functionify some of the main loop, so it isn't so horrificly deep and isBob Beck
a bit easier to look at, for small values of easier. -Add two options for -u user and -g group to optionally make the proxy drop privs after doing it's pf ioctl's to find out where to go. Running as non root does mean that the PORT and EPRT backchannels do not come from port 20, but this isn't a problem for most sensible ftp clients and sets of packet filter rules that aren't written by a knuckle dragging ape living in the 90's. I would make it drop privs by default, but technically this breaks the ftp specs, and for the upcoming stuff to deal with EPRT, we will need root privs to manipulate rdr rules).
2001-08-21Fix up comment to indicate order that libraries in which theDale Rahn
libraries are actually searched.
2001-08-20Powered by @mantoya.Mike Pechkin
o) we don't like .Nd without argument, except next line starts with a new tag; millert@ ok
2001-08-20Powered by @mantoya.Mike Pechkin
o) fix bogus .Xr usage; millert@ ok.
2001-08-19unsmokecrack - thanks millertBob Beck
2001-08-19decruftify with prejudiceBob Beck
2001-08-19decruftifyBob Beck
2001-08-19Argh. I must be wearing a stupid sign todayBob Beck
2001-08-19bye bye atoi.Bob Beck
2001-08-19oopsBob Beck