Skip to content
This repository has been archived by the owner on Dec 1, 2020. It is now read-only.

Check memory leak in nuttx #6

Closed
seanshpark opened this issue Sep 16, 2015 · 3 comments
Closed

Check memory leak in nuttx #6

seanshpark opened this issue Sep 16, 2015 · 3 comments

Comments

@seanshpark
Copy link
Contributor

tuvtester running in nuttx.

First run

nsh> free
             total       used       free    largest
Mem:        165104      13136     151968      97376

Second run

nsh> free
             total       used       free    largest
Mem:        165104      13472     151632      97376
@seanshpark
Copy link
Contributor Author

Testing with valgrind / linux reports no leaks.

@seanshpark
Copy link
Contributor Author

When pipe() is called /dev/pipe%d device is created with register_driver() for the pipe but is not removed when both handles are closed. As device can be reused it is OK not to unregister for performance. But it's confusing whether there is other memory leak or not.

seanshpark added a commit that referenced this issue Sep 30, 2015
* related with #6

libtuv-DCO-1.0-Signed-off-by: SaeHie Park [email protected]
@seanshpark
Copy link
Contributor Author

Memory leak is fixed for linux and also for nuttx. There was one for linux when checked with valgrind.
For nuttx, as /dev/pipe? file allocates memory, should have to delete the file to check the leak.
But as there is some problem with g_pipeset in nuttx/drivers/pipes/pipe.c sync with pipecommon_freedev() call, second run of tuvtester fails.

tadziopazur pushed a commit to tadziopazur/libtuv that referenced this issue Oct 19, 2017
Recently, we reduced the binary size of http-parser (PR Samsung#6 in http-parser).
This patch enables this feature.

IoT.js-DCO-1.0-Signed-off-by: Sanggyu Lee [email protected]
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant