Age | Commit message (Collapse) | Author | |
---|---|---|---|
2007-01-14 | move things around in rcs_rev_getbuf() and rcs_rev_write_fd() | Joris Vink | |
so that we do keyword expansion on-the-fly if required instead of obtaining the revision in memory first, running over the revision lines while expanding keywords and only then writing them to the fd or memory buffer. this drasticly decreases cpu usage by opencvs on very large trees (like src). OK niallo@ | |||
2007-01-14 | reset sticky tags with -A; OK joris@. | Xavier Santolaria | |
2007-01-14 | replace checkout_write_revision() with rcs_rev_write_stmp() | Joris Vink | |
they do exactly the same anyway. | |||
2007-01-13 | nuke the the very evil rcs_getrev() stuff | Joris Vink | |
discussed with niallo@ | |||
2007-01-13 | make import and init work again after the small breakage | Joris Vink | |
we had with the rcs changes; | |||
2007-01-13 | switch commit code to the new RCS api. | Joris Vink | |
kay niallo@ | |||
2007-01-13 | make a copy of cf->file_ent->ce_rev in cf->file_rcsrev when needed | Joris Vink | |
otherwise we might get a NULL dereference when free'ing the cf struct with cvs_file_free() | |||
2007-01-13 | plug 3 memleaks | Joris Vink | |
2007-01-13 | remove BUF * argument from cvs_checkout_file() it will | Joris Vink | |
no longer be used, ever. | |||
2007-01-13 | - remote update now works again, with new rcs api | Joris Vink | |
- we no longer hold the entire file in memory when sending or receiving it | |||
2007-01-13 | the memleak i thought i had fixed was in fact NOT a leak at all | Joris Vink | |
and caused opencvs to crash so revert, oops? | |||
2007-01-13 | - better error messages | Joris Vink | |
- correclty use cvs_tagname, so that the tag from $CVSROOT/config will actually be expanded as a keyword. | |||
2007-01-13 | add missing prototypes for some new rcs api functions; | Joris Vink | |
2007-01-12 | simplify after recent rcs API changes. | Joris Vink | |
kay niallo@ | |||
2007-01-12 | major re-work of the RCS api. this results in 100x performance improvements ↵ | Niall O'Higgins | |
in some places and much reduced memory usage. note that only checkout has been fully converted to use the new high-performance functions. other codepaths (e.g. update) still use the old method which is provided for backwards compatibility. we can convert the remaining slow bits over to the new api piece-by-piece. "commit this, now" joris@ brad@ tested, too. | |||
2007-01-12 | more secsh -> rfc 4716 updates; | Jason McIntyre | |
spotted by wiz@netbsd ok markus | |||
2007-01-12 | do not leak memory from rcs_head_get() | Joris Vink | |
ok niallo@ | |||
2007-01-12 | if we have no files to be commited, stop commiting. | Joris Vink | |
2007-01-12 | and thou shalt not leak memory when removing entries | Joris Vink | |
2007-01-12 | complete binary support for opencvs, | Joris Vink | |
adding/importing binary files now works too partially taken from openrcs | |||
2007-01-12 | link to build the edit, editors, unedit, watch, and watchers commands. | Xavier Santolaria | |
more work needs to be done for them though. | |||
2007-01-12 | fix incorrect output which was driving me more then crazy | Joris Vink | |
2007-01-11 | missed a bit. pointed out by jmc@ | Niall O'Higgins | |
2007-01-11 | sync a bit with reality | Joris Vink | |
2007-01-11 | remove some un-needed buffer NUL-termination. in openrcs, diff3 is already ↵ | Niall O'Higgins | |
NUL-terminating its buffers so we don't need to do it twice. this allows us to sync rcs_splitlines() with opencvs' cvs_splitlines(). ok joris@ | |||
2007-01-11 | silence some warnings. | Jasper Lievisse Adriaanse | |
ok joris@ | |||
2007-01-11 | rework opencvs so that we can deal with binary files. previously we assumed ↵ | Niall O'Higgins | |
all files were ascii, which broke things in real-world usage. now a checkout of src should work, albeit using lots of memory and cpu. fixing this is the next step. testing by many. ok & some input joris@ | |||
2007-01-11 | update the revision number in CVS/Entries from CVS/Baserev if we | Xavier Santolaria | |
are reverting to the revision that the user edited in the first place. ok joris@. | |||
2007-01-11 | revert last commit after discussing w/ joris@; makes sense to remove | Xavier Santolaria | |
the tmp files once we'r done w/ them, not at the very end only.. | |||
2007-01-11 | no need to unlink() the tmp files here as they have already been added | Xavier Santolaria | |
to the temp_files walklist anyway.. | |||
2007-01-11 | strl* overload, use xasprintf(). | Xavier Santolaria | |
2007-01-11 | strl* overload, use xasprintf(). | Xavier Santolaria | |
2007-01-11 | save errno. | Xavier Santolaria | |
2007-01-11 | sync joris' last change (connection to remote server when all options are | Xavier Santolaria | |
valid) for the unedit command too. | |||
2007-01-11 | in a remote setup: | Joris Vink | |
do not connect to the remote server until we are sure all the options passed to the commands are valid. noticed by xsa@ | |||
2007-01-10 | add a cvs_file_copy routine and use it for the edit command to | Xavier Santolaria | |
copy the file in the current working directory to CVS/Base/ ok ray@ joris@. | |||
2007-01-10 | you know what, now that openrcs is in fact working properly and | Joris Vink | |
has been in the previous release - how about we give it a proper version instead of "OpenCVS RCS 3.6" okay xsa@, ray@ | |||
2007-01-10 | - use .Sq for single characters, rather than .Dq | Jason McIntyre | |
- zap trailing whitespace | |||
2007-01-10 | do not use a list for SYNOPSIS; | Jason McIntyre | |
this is actually part of a larger report sent by eric s. raymond and forwarded by brad, but i only read half of it. spotted by brad. | |||
2007-01-10 | Clear EOF before prompting user for input. | Ray Lai | |
Fixes PR 5342. OK joris@ and xsa@. | |||
2007-01-09 | fix usage | Theo de Raadt | |
2007-01-09 | add remote handler for the cvs watch command; not linked to the build. | Xavier Santolaria | |
2007-01-09 | Ignore "mapping symbols" like $a and $t. They seem to be an ARM-only | Chris Kuethe | |
feature, used to indicate whether code is ARM or Thumb. Unfortunately they confuse gprof, which outputs call graphs where every other function is named "$a"... not very useful. Rather than enumerating the different symbols, binutils ignores anything beginning with '$', and that is what we will do here. Thanks to Dale Rahn for useful tips along the way. ok miod | |||
2007-01-09 | remove uneeded linebreak when creating the CVS/Notify | Xavier Santolaria | |
file with the edit command. | |||
2007-01-09 | Accept empty command, from Matthew R. Dempsky; ok deraadt@ millert@ | Otto Moerbeek | |
2007-01-08 | - simplify CVS/Baserev file parsing; makes unedit work better. | Xavier Santolaria | |
- add tracing messages in edit, unedit functions. | |||
2007-01-07 | add the temporary file we create to the temp_files worklist | Joris Vink | |
2007-01-07 | make the output when interrupting writing the log message consistent. | Jasper Lievisse Adriaanse | |
ok joris@ commiting with opencvs now btw :) | |||
2007-01-07 | use cvs_logmsg_read() for -F for simplicity. | Joris Vink | |
2007-01-07 | and lets make import work with the new logmsg code too! | Joris Vink | |