You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This benchmark requires running make in the directory to build the addon before it can be run (also the Makefile tries to use the global node-gyp, not the one in the codebase). Currently running it with run.js it will just report misc/function_call.js Binding failed to load.
Should we keep this benchmark? If so, it needs to be fixed to work with our current benchmark suite. The purpose of this benchmark seems to be:
// show the difference between calling a short js function
// relative to a comparable C++ function.
// Reports millions of calls per second.
// Note that JS speed goes up, while cxx speed stays about the same.
sounds useful to me.
The text was updated successfully, but these errors were encountered:
I think you are supposed to invoke the benchmarks through make bench or make bench-misc, that will build the add-on. There is probably room for improvement, though.
Move the logic for building the benchmark/misc/function_call to
the top-level Makefile and use our local copy of node-gyp.
PR-URL: nodejs/node#16160Fixes: nodejs/node#16154
Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
Reviewed-By: Franziska Hinkelmann <franziska.hinkelmann@gmail.com>
Move the logic for building the benchmark/misc/function_call to
the top-level Makefile and use our local copy of node-gyp.
PR-URL: #16160Fixes: #16154
Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
Reviewed-By: Franziska Hinkelmann <franziska.hinkelmann@gmail.com>
This benchmark requires running
make
in the directory to build the addon before it can be run (also theMakefile
tries to use the globalnode-gyp
, not the one in the codebase). Currently running it withrun.js
it will just reportmisc/function_call.js Binding failed to load
.Should we keep this benchmark? If so, it needs to be fixed to work with our current benchmark suite. The purpose of this benchmark seems to be:
sounds useful to me.
The text was updated successfully, but these errors were encountered: