diff options
| author | Florian Fischer <florian.fl.fischer@fau.de> | 2019-08-11 22:28:38 +0200 |
|---|---|---|
| committer | Florian Fischer <florian.fl.fischer@fau.de> | 2019-08-11 22:40:06 +0200 |
| commit | 9d056ff64bf887e2c3479fe515bf7c95cb39e5b6 (patch) | |
| tree | d9c2d9819e9b0a175721562db546fbc25f3ca54f /src/run_cmd.c | |
| parent | f7d64e02af941515bc894dea10f55255d06cbcf7 (diff) | |
| download | allocbench-9d056ff64bf887e2c3479fe515bf7c95cb39e5b6.tar.gz allocbench-9d056ff64bf887e2c3479fe515bf7c95cb39e5b6.zip | |
Rework exec chain
Originally the structure of the executed cmd was
{measure cmd} {allocator cmd prefix} {cmd} with the parent environment
except LD_PRELOAD was modified for the whole command chain.
Unfortunatly perf causes segfaults with some allocators and measuring
allocators cmd prefixes doesnt seem fair. So the new cmd chain looks like:
{allocator cmd prefix} {measure cmd} run_cmd <LD_PRELOAD> {cmd} without
touching the environment in python.
run_cmd sets LD_PRELOAD to the value it received in argv[1] and executes
argv[2] with the rest of argv. This does also measure code not part of
the actual benchmark but in a equal manner and not only for some allocators.
Diffstat (limited to 'src/run_cmd.c')
| -rw-r--r-- | src/run_cmd.c | 18 |
1 files changed, 18 insertions, 0 deletions
diff --git a/src/run_cmd.c b/src/run_cmd.c new file mode 100644 index 0000000..aa15964 --- /dev/null +++ b/src/run_cmd.c @@ -0,0 +1,18 @@ +#define _GNU_SOURCE +#include <stdio.h> +#include <stdlib.h> +#include <unistd.h> + +int main(int argc, char* argv[]) { + if (argc < 3) { + printf("Usage: %s <ld_preload> <cmd> [cmd args]\n"); + printf("\tset LD_PRELOAD to ld_preload and call execvp <cmd> [cmd args]\n"); + return 1; + } + + // Overwrite LD_PRELOAD. + setenv("LD_PRELOAD", argv[1], 1); + + // Run cmd. + execvp(argv[2], &argv[2]); +} |
