OS_ModuleUnload() for statically loaded module may segfault #637
Comments
jphickey
added a commit
to jphickey/osal
that referenced
this issue
Oct 30, 2020
Ensure that the handle is not NULL before invoking dlclose(). In particular the handle will be NULL for static modules.
jphickey
added a commit
to jphickey/osal
that referenced
this issue
Nov 2, 2020
Mark static modules with a flag, and check that flag at unload. If flag is set, then skip unload low level implementation.
jphickey
added a commit
to jphickey/osal
that referenced
this issue
Nov 4, 2020
Mark static modules with a different type, and check that type at unload. Only call unload low level implementation if type is dynamic.
astrogeco
added a commit
that referenced
this issue
Nov 10, 2020
Fix #637, fix OS_ModuleUnload for static modules
jphickey
pushed a commit
to jphickey/osal
that referenced
this issue
Aug 10, 2022
jphickey
pushed a commit
to jphickey/osal
that referenced
this issue
Aug 10, 2022
Fix nasa#637, Remove old CFE_ARINC653-related tests
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Describe the bug
Calling
OS_ModuleUnload()on a statically loaded module will still attempt to calldlclose()but the handle is NULL because no module was loaded. At least on Linux/Glibc this actually causes a segmentation fault.To Reproduce
Using the CFE framework switch any app (e.g. sample_app) from a regular dynamic app to a static app.
Build and run CFE, then shutdown with CTRL+C.
The shutdown ends up calling
OS_DeleteAllObjects()which in turn will trigger this bug - one gets a segfault instead of a clean exit.Expected behavior
Should shutdown cleanly.
System observed on:
Ubuntu 20.04
Reporter Info
Joseph Hickey, Vantage Systems, Inc.
The text was updated successfully, but these errors were encountered: