aboutsummaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* add el_new and el_destroy functions for stable ABIstable-abiMichał Łyszczek2019-05-1412-4/+234
| | | | | | | | | | | | | | | | | | | | | | el_new will allocate on heap "el" object, so user does not have to know any internals of "struct el". This will make ABI stable in case when anything in "struct el" changes. el_oinit can still be used to allocate memory on stack, but with it ABI may be broken, so it should be used only on systems when user can afford to recompile all programs that use embedlog lib. * include/embedlog.h.in add 2 new functions, el_init and el_destroy * man/ add description for 2 new functions, describe notes about stable ABI and stack allocated methods. * tst/ add test for 2 new functions Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* API change: rename "struct el_options" to "struct el"Michał Łyszczek2019-05-1424-658/+661
| | | | Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* include: create embedlog.h.in and generate embedlog.h with itMichał Łyszczek2019-05-142-0/+3
| | | | Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* src/el-options: use static value of 0x7fff for EL_OUT_ALLMichał Łyszczek2019-05-134-4/+35
| | | | | | | | | | | | | | | | | | | | | | | | | | EL_OUT_ALL now performs purely abstract role and is not used to check for valid errors - there is new ALL_OUTS define for that in el-private.h. This change is needed so ABI does not change when we add new output. * include/embedlog.h change EL_OUT_ALL value from dynamically calculated mask to static 0x7fff. Previously it was mask of all implemented outputs, now its role is pure abstract. * src/el-options.c when abstract EL_OUT_ALL value is passed to EL_OUT option, program will treat it as if VALID_OUTS was passed. EINVAL is returned now based on new ALL_OUTS define instead of EL_OUT_ALL. * src/el-private.h this is where ALL_OUTS is defined * tst/test-el-options.c update tests for the change Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* include/embedlog.h: add static values to enumsMichał Łyszczek2019-05-132-44/+68
| | | | | | | | | Adding new enums in place different than at the very end of list results in ABI breakage. Adding static values to enum ensures that ABI stays stable even in event of sorting enums (to increase readability). Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* el_flush(): add new function to force data from buffers to deviceinnogyMichał Łyszczek2019-03-2916-69/+318
| | | | Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* src/el_options.c: add function to get global options objectMichał Łyszczek2019-03-285-0/+42
| | | | | | | That object can be used to read embedlog options - for example to clone printing settings. Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* embedlog.h: add user's pointer to el_custom_putsMichał Łyszczek2019-03-277-14/+25
| | | | | | | | el_custom_puts now also accepts "void *user" pointer beside standard "const char *s". It's usefull when you need to pass pointer to logger object that is not global. Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* man/el_option.3: add missing EL_FUNCINFO optionMichał Łyszczek2019-03-021-0/+17
| | | | | no_buildbot Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* man/el_print.3: el_pmemory: desribe case when mlen == 0Michał Łyszczek2019-03-021-0/+5
| | | | | | | spoiler alert: nothing happens no_buildbot Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* ABI/API change: add printing function name with logsMichał Łyszczek2019-03-0114-130/+445
| | | | | | only on c99+ compilers Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* version bump to v0.5.0Michał Łyszczek2019-02-287-7/+7
|
* API change: rename EL_FILE_SYNC_* to EL_FSYNC_* to be consistentMichał Łyszczek2019-02-236-32/+32
| | | | | | | | | I have *no* idead why I called it _FILE_. Must be one of those youth mistakes. This doesn't affect ABI. Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
* fix: month language in man pages, doh...v0.4.0Michał Łyszczek2019-02-155-5/+5
| | | | no_buildbot
* fix: some package building problemsMichał Łyszczek2019-02-154-9/+9
| | | | no_buildbot
* remove some code when binary logs are disabledMichał Łyszczek2019-02-132-9/+16
| | | | | | This removes code that was comiled regardless of options but that code makes sense only when binary logs are enabled. Now that code is enabled only when binary logs are enabled. You can never save enough memory.
* src/el-private.h: remove all existance of feature test macroMichał Łyszczek2019-02-131-116/+0
| | | | | | | | This. Was. BAD. Idea. Seriously, if you ever thing about defining these again - think again. Then again, and again. If still you want to use them - this once again. It's so much pain in the ass and little gain.
* src/el-pmemory.c: fix usage of uninitialized variableMichał Łyszczek2019-02-131-0/+2
|
* src/el-pmemory.c: fix wrong options object usedMichał Łyszczek2019-02-131-1/+1
|
* bump version of manual pagesMichał Łyszczek2019-02-135-5/+5
|
* add: make coverage instructionsMichał Łyszczek2019-02-133-0/+31
|
* some visual tweaks to manual pagesMichał Łyszczek2019-02-139-27/+45
|
* add: option to print to stdoutMichał Łyszczek2018-12-057-11/+37
|
* options: fix g_options altered after el_ocleanupMichał Łyszczek2018-11-242-1/+53
| | | | | | calling el_ocleanup() changed outputs in global g_options object, rendering all el_print function family to stop printing after any el_ocleanup() call.
* make embedlog print to stderr by defaultMichał Łyszczek2018-11-238-22/+8
|
* bump version for next releaseMichał Łyszczek2018-11-232-2/+2
|
* man/el_print.3: fix typo in encoded integer example tableMichał Łyszczek2018-11-231-1/+1
|
* add: printing memory without ascii tableMichał Łyszczek2018-11-2311-89/+214
| | | | | That ascii table could take a lot of space, especially when printing single line, now it's an option
* pkg/deb: fix description in control templateMichał Łyszczek2018-11-231-6/+6
| | | | no_buildbot
* www: update index and download pagesMichał Łyszczek2018-11-152-7/+182
| | | | no_buildbot
* pkg/tgz: fix wrong conditional when runnign testprogMichał Łyszczek2018-11-141-1/+1
| | | | no_buildbot
* pkg/tgz: add CFLAGS and LDFLAGS for building tgz packageMichał Łyszczek2018-11-141-1/+1
| | | | no_buildbot
* pkg/*: test if testprog fails when there is no lib in the system anymoreMichał Łyszczek2018-11-143-0/+12
| | | | no_buildbot
* man: fix typo in macroMichał Łyszczek2018-11-141-1/+1
| | | | no_buildbot
* pkg/deb: add CFLAGS and LDFLAGS for debuildMichał Łyszczek2018-11-141-0/+6
| | | | no_buildbot
* pkg/deb: add missing install filesMichał Łyszczek2018-11-143-1/+6
| | | | no_buildbot
* pkg/deb: change generated package name from embedlog0 to libembedlog0Michał Łyszczek2018-11-144-18/+17
| | | | | | | debian requires that packages containing libraries be named just like lib in /usr/lib no_buildbot
* pkg/deb: use @{ instead of ${ as replace tokenMichał Łyszczek2018-11-141-2/+2
| | | | no_buildbot
* makefile: do not build examples with make allMichał Łyszczek2018-11-142-2/+14
| | | | no_buildbot
* pkg: add basic files to create packagesMichał Łyszczek2018-11-1412-0/+481
| | | | no_buildbot
* add: feature test macros for ONLCR for different systemsMichał Łyszczek2018-10-121-2/+16
|
* clean that mess with feature test macro once and for allfeature-test-macro-mess-fixMichał Łyszczek2018-10-1221-138/+120
|
* change \e to \033 in testsMichał Łyszczek2018-10-121-18/+18
|
* fix: compilation error with bad feature test macro on some systemsMichał Łyszczek2018-10-124-29/+59
| | | | some systems defines features in different posix versions
* fix: compilation error on solarisMichał Łyszczek2018-10-121-0/+18
| | | | | | solaris will crap out when _XOPEN_SOURCE is set to 500, and you try to compile it with c99 compiler, stupid, but for solaris force -std=c89 flag
* ok, I suppose defining feature test macro locally is better ideaMichał Łyszczek2018-10-113-6/+21
| | | | it caused some systems to fail to compile it
* make code more c89 compatibleMichał Łyszczek2018-10-114-28/+64
|
* fix: remove comma at the end of enum listMichał Łyszczek2018-10-111-1/+1
|
* remove check for memcpy, it is useless and breaks clang sanitizerMichał Łyszczek2018-10-111-2/+0
| | | | that is bug in clang.
* fix: add missing #ifdef LLONG_MAX in binary testsMichał Łyszczek2018-10-101-0/+13
| | | | | This caused error failure on machines where long long is not available