aboutsummaryrefslogtreecommitdiff
path: root/src/nvim/profile.c
Commit message (Collapse)AuthorAge
* chore: use codespell to spell check #15016dundargoc2021-07-07
|
* reltimefloat(): allow negative result #10544Justin M. Keyes2019-07-20
| | | | | | | | | For "backwards" duration, reltimefloat() should return negative value like its counterpart reltimestr(). ref bab24a88ab48 ref 06af88cd72ea ref #10521 fix #10452
* viml/profile: revert proftime_T to unsigned type #10521Justin M. Keyes2019-07-16
| | | | | | | | | | | | | | - reltimestr(): Produce negative value by comparing the unsigned proftime_T value to INT64_MAX. https://github.com/neovim/neovim/issues/10452#issuecomment-511155132 1. The interfaces of nearly all platforms return uint64_t. INT64_MAX is only half of that. 2. Low-level interfaces like this typically define that there is no fixed starting point. The only guarantees are that it's (a) monotonically increasing at a rate that (b) matches real time. ref 06af88cd72ea fix #10452
* viml/profile: cast os_hrtime() resultJustin M. Keyes2019-07-16
| | | | | | | | | | | | | | | | Temporary measure to avoid QuickBuild CI failure: 8:42:54,702 INFO - Executing post-execute action... 18:42:54,702 ERROR - Step 'master>buildall>build-node?testNode=freebsd-64>build-and-run-tests>build-and-run-tests-parameterized?buildType=Release>configure-neovim-and-build-nvim' is failed: Failed to run command: mkdir -p build/Release && cd build/Release && cmake -G "Unix Makefiles" -DBUSTED_OUTPUT_TYPE=TAP -DMIN_LOG_LEVEL=3 -DCMAKE_BUILD_TYPE=Release -DTRAVIS_CI_BUILD=ON ../.. && gmake VERBOSE=1 nvim unittest-prereqs functionaltest-prereqs Command return code: 2 Command error output: /usr/home/quickbuild/buildagent/workspace/root/neovim/pull-requests-automated/src/nvim/profile.c:70:27: error: implicit conversion changes signedness: 'proftime_T' (aka 'long') to 'unsigned long' [-Werror,-Wsign-conversion] STRICT_ADD(os_hrtime(), nsec, &rv, int64_t); ~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~ /usr/home/quickbuild/buildagent/workspace/root/neovim/pull-requests-automated/src/nvim/assert.h:150:26: note: expanded from macro 'STRICT_ADD' do { *(c) = (t)((a) + (b)); } while (0) ^ 1 error generated. gma...
* viml/profile: revert gettimeofday() #10488Justin M. Keyes2019-07-13
| | | | | | | e2ce5ff9d616 was proven to be bogus, so revert it. close #10328 ref #10356 ref #10452
* viml/reltime(): allow negative result #10453Justin M. Keyes2019-07-09
| | | | | | - define proftime_T as signed integer - profile_sub(): allow negative result closes #10452
* viml/profile: switch to uv_gettimeofday() #10356Justin M. Keyes2019-06-29
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Performance of high-resolution time (clock_gettime via uv_hrtime) is expensive on some systems. For profiling VimL, syntax, etc., we don't care about nanosecond-precision and monotonicity edge-cases, so avoid uv_hrtime(). closes #10328 From the uv__hrtime() source: https://github.com/libuv/libuv/blob/0cdb4a5b4b706d0e09413d9270da28f9a88dc083/src/unix/linux-core.c#L442-L462 /* Prefer CLOCK_MONOTONIC_COARSE if available but only when it has * millisecond granularity or better. CLOCK_MONOTONIC_COARSE is * serviced entirely from the vDSO, whereas CLOCK_MONOTONIC may * decide to make a costly system call. */ This micro-benchmark (Debug build) shows negligible differences on my system: #include <sys/time.h> ... proftime_T tm = profile_start(); int trials = 999999; int64_t t = 0; struct timeval tv; for (int i = 0; i < trials; i++) { t += gettimeofday(&tv,NULL); } tm = profile_end(tm); ILOG("%d trials of gettimeofday: %s", trials, profile_msg(tm)); tm = profile_start(); for (int i = 0; i < trials; i++) { t += os_hrtime(); } tm = profile_end(tm); ILOG("%d trials of os_hrtime: %s", trials, profile_msg(tm)); tm = profile_start(); for (int i = 0; i < trials; i++) { t += os_utime(); } tm = profile_end(tm); ILOG("%d trials of os_utime: %s", trials, profile_msg(tm)); ILOG("%zu", t);
* *: Add comment to all C filesZyX2017-04-19
|
* MSVC: include os/os_defs.h for `inline`. #3239Rui Abreu Ferreira2015-08-25
|
* startuptime: always enable startuptimeNicolas Hillegeer2014-07-20
| | | | Removes the STARTUPTIME define.
* startuptime: implement on top of profiling fnsNicolas Hillegeer2014-07-20
| | | | | | Reuse the profiling functions to implement the startuptime functions. Decreases our dependency on `gettimeofday()` and thus gets us a little bit closer to a clean port to Windows.
* startuptime: move code to profile.{c,h} + docNicolas Hillegeer2014-07-20
| | | | | It's a better place to put it. Also slightly documented and reformatted, but not changed.
* profiling: implement on top of os_hrtime()Nicolas Hillegeer2014-07-16
Should be better than gettimeofday() since libuv uses higher resolution clocks on most UNIX platforms. Libuv also tries to use monotonic clocks, kernel bugs notwithstanding, which is another win over gettimeofday(). Necessary for Windows, which doesn't have gettimeofday(). In vanilla vim, Windows uses QueryPerformanceCounter, which is the correct primitive for this sort of things, but that was removed when slimming up the codebase. Libuv uses QueryPerformanceCounter to implement uv_hrtime() on Windows so the behaviour of vim profiling on Windows should now be the same. The behaviour on Linux should be different (better) though, libuv uses more accurate primitives than gettimeofday(). Other misc. changes: - Added function attributes where relevant (const, pure, ...) - Convert functions to receive scalars: Now that proftime_T is always a (uint64_t) scalar (and not a struct), it's clearer to convert the functions to receive it as such instead of a pointer to a scalar. - Extract profiling funcs to profile.c: make everything clearer and reduces the size of the "catch-all" ex_cmds2.c - Add profile.{c,h} to clint and -Wconv: - Don't use sprintf, use snprintf - Don't use long, use int16_t/int32_t/...