anchao commented on PR #17468:
URL: https://github.com/apache/nuttx/pull/17468#issuecomment-3832940464

   > > > > I won't let this submission be merged:
   > > > > 
   > > > > 1. code size will increase.
   > > > > 2. performance will degrade.
   > > > 
   > > > 
   > > > without refcount, the used after free will happen under the stress 
testing of task/thread create/destroy. Used after free is a critical bug which 
must be fixed. refcount is one of well known method to fix this type of error, 
if you have better suggestion/method, please point out.
   > > 
   > > 
   > > 
   > > 1. This is because in your business logic, Task A still holds the 
resources of Task B when it exits. Why can't the resources be released properly?
   > > 2. Most IoT and wearable devices use a fixed thread pool. If your 
business process involves the frequent creation and destruction of a large 
number of threads, why not design them as deterministic tasks and implement the 
creation and exit logic with wait/notify mechanisms?
   > 
   > No, I just consider POSIX spec, not private business logic. POSIX spec 
allow appliation create/destroy task/thread dynamically, we need fix the 
problem. Let's hightlight the inviolable rule here: 
https://github.com/apache/nuttx/blob/master/INVIOLABLES.md#strict-posix-compliance
   > 
   > > 3. The only acceptable solution to me is to make this feature 
configurable via Kconfig – no other options are on the table.
   > 
   > it's bug fix, I don't unerstand why we add an option to choice whether to 
fix a critical(memory corruption) bug.
   
   But this is not a bug on my end. There is no scenario in my commercial 
solution where threads are dynamically created or deleted. If you insist on 
merging this commit, make it configurable—this way it won’t block existing 
users. This is my final concession.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to