Age | Commit message (Collapse) | Author | |
---|---|---|---|
2001-06-13 | print the name | Theo de Raadt | |
2001-06-13 | Add support for lazy binding. | Artur Grabowski | |
For now we only do it for the main executable, some code cleanup is necessary to support it for libraries. This fixes the problem I had in cc1plus and I'm able to finish a make build with shared libraries enabled. | |||
2001-06-13 | Lazy binding entry point. From NetBSD with some help from FreeBSD. | Artur Grabowski | |
2001-06-13 | Don't do plt relocations here, move that to md code that might choose | Artur Grabowski | |
to do lazy binding. | |||
2001-06-08 | Cleanup. | Artur Grabowski | |
exit when the bootstrap relocation fails. | |||
2001-06-08 | No more need to pull in strlen from libc. | Artur Grabowski | |
2001-06-08 | Be consistent. If we have inveted our of strlen, use it everywhere. | Artur Grabowski | |
2001-06-08 | style | Artur Grabowski | |
2001-06-08 | KNF | Theo de Raadt | |
2001-06-07 | It started with trying to write code to avoid running static binaries, | Artur Grabowski | |
it ended with a complete rewrite. XXX - Doesn't handle '-x' yet, but the old didn't do that either (and didn't even match the man page). | |||
2001-06-06 | more cleaning | Artur Grabowski | |
2001-06-06 | cleaning | Artur Grabowski | |
2001-06-06 | Replace all 'if (_dl_debug) _dl_printf(...)' with a macro. | Artur Grabowski | |
2001-06-06 | No need to extern _dl_debug here. | Artur Grabowski | |
2001-06-06 | more cleanup | Artur Grabowski | |
2001-06-06 | Cleanup. | Artur Grabowski | |
- remove printfs that are only useful for the person who put them there. - Style. - dead code. Needs much more work. | |||
2001-06-05 | gcc generates a lot of unaligned RELATIVE relocations for c++ exception ↵ | Artur Grabowski | |
handlers. Workaround that. | |||
2001-06-01 | - Allocate space for the finishing NUL in the name of ld.so. | Artur Grabowski | |
- some comments. | |||
2001-05-31 | Apparently GLOB_DAT needs r_addend. | Artur Grabowski | |
"Hello, world!\n" now works with libc.so | |||
2001-05-31 | Some cleanups. | Artur Grabowski | |
2001-05-31 | unsigned long, not u_int32_t | Artur Grabowski | |
2001-05-31 | If we can't find atexit (because we linked against libc.a), | Artur Grabowski | |
print a message instead of dereferencing a NULL pointer. | |||
2001-05-31 | alpha needs JMPREL relocations. | Artur Grabowski | |
various cleanups. | |||
2001-05-31 | Remove write protection from some memory segments when | Artur Grabowski | |
doing rel and rela relocations. XXX - we have to find out when and why we actually need it. | |||
2001-05-31 | random indentation fixes (needs much more work). | Artur Grabowski | |
2001-05-31 | Correctly implement most linking for alpha. | Artur Grabowski | |
Remove a lot of code copied from powerpc. | |||
2001-05-31 | When opening the hints file, check for error return < 0, not -1. | Artur Grabowski | |
_dl_open is not necessarily like the regular open(2). XXX - maybe we should check with _dl_check_error? | |||
2001-05-29 | Check for error on open. | Artur Grabowski | |
ENOENT happens to be the same as stderr. :) | |||
2001-05-28 | Commonize csu code for elf systems, powerpc now no longer has it's own | Dale Rahn | |
versions of these files. Fixed a bug in ld.so in this, instead of scheduling the fini of each of the shared libraries with atexit. schedule a function of ld.so itself and it will walk all of the open libraries when the program exits. otherwise a shared library could be dl_open()ed and then dl_close()d and then it would not be mapped for the atexit processing. TODO: What if atexit is not found (process did not link against libc?) Do shared libraries that are dl_closed have their global destructors run? | |||
2001-05-23 | Do not enable debugging, such as PRINTF_DEBUG for ld.so, it becomes | Dale Rahn | |
too verbose for normal usage. | |||
2001-05-16 | proper typing of two variables | Theo de Raadt | |
2001-05-15 | Niklas, there is no excuse for breaking this code | Theo de Raadt | |
2001-05-14 | Commit initial alpha bits, to facilitate in-tree dev. | Niklas Hallqvist | |
This code is not yet working. | |||
2001-05-12 | MAP_COPY -> MAP_PRIVATE | Artur Grabowski | |
2001-05-11 | -1 -> MAP_FAILED. | Artur Grabowski | |
2001-05-11 | MAP_COPY -> MAP_PRIVATE | Artur Grabowski | |
2001-05-11 | -1 -> MAP_FAILED | Artur Grabowski | |
2001-05-11 | mmap returns MAP_FAILED when it fails, not -1 | Artur Grabowski | |
2001-05-11 | MAP_COPY -> MAP_PRIVATE | Artur Grabowski | |
2001-04-11 | Correct .Nm usage. | Aaron Campbell | |
2001-04-02 | Cleanup for 64bit support. | Dale Rahn | |
Pieces by art, niklas and me. Only tested on powerpc. | |||
2001-03-30 | Add infrastructure to allow mapping of text sections which are normally | Dale Rahn | |
RO, RW while ld.so is working. And then the information to set the sections back to RO (or appropriate mode). PowerPC now supports the typical NON-PIC relocations in ld.so. I do not know how well this will work with large shared libraries. I seem to recall a possible problem with large data where data is located in a different shared library. | |||
2001-02-16 | Now that powerpc is using UVM, this shared library hack is no longer necessary. | Dale Rahn | |
2001-02-03 | Previously only powerpc used the DT_INIT field to find the ctors/init | Dale Rahn | |
code, now that will be changed to all arch besides mips will default to using that. mips probably would be changed to but it has not been tested. | |||
2001-01-30 | sync with a.out ldconfig, adds missing -U option. Reviewed by: drahn@ | Brad Smith | |
2001-01-28 | $OpenBSD$ | Niklas Hallqvist | |
2001-01-23 | fix includes | Theo de Raadt | |
2000-11-09 | Change all option list specifications to ".Bl -tag -width Ds". Most man | Aaron Campbell | |
pages just needed their -width parameter tweaked to "Ds", which provides a nice width of 6 constant characters. For consistency more than anything. | |||
2000-10-30 | libexec man page fleshing. again, bored on the plane home. | Aaron Campbell | |
2000-10-19 | Fix bug in dl_strcmp, it is not always legal to access a pointer | Dale Rahn | |
at offset -1, the code that was here was very odd, the corrected code was basically copied from the libsa version. |