Xiang Xiao

@tito97sp could you squash the intermediate change into one patch?

look like the patch become to empty and confuse the github

@tito97sp will you submit a new PR?

@paolovolpi could you provide a patch?

it can be added in the separate PR. Maybe need to enabled in the debug mode to save the space.

Yes, this is the safe path to fix the priority inheritance issue.

> One of the possible ways may be to get back original code for the files that I've listed in this PR and switching `nxmutex_` APIs to use `nxsem_` APIs...

> The issue of naming conflict comes from: `#include ` -> `#include ` -> `#include ` because `struct file_struct` has member `rmutex_t fs_lock;` It's hard to fix:(

Here is the reason: 1. All l2cc_xxx hold the critical section internally 2. All other up_xxx functions in the same file doesn't hold the critical section

> I do not have a strong opinion on this. The PR description states "since l2cc_invalidate_all already do the protection" ("callee" direction), but discussion goes into "caller" direction, so that...