Issue created by Sebastian Huber: 
https://gitlab.rtems.org/rtems/rtos/rtems/-/work_items/5706



`rtems_rap_load()` and `rtems_rap_unload()` take the RAP lock and then call
`rtems_rap_find()`, which takes it again. The lock is an ordinary `rtems_mutex`,
so the second acquisition is a self-deadlock and the system terminates with
`INTERNAL_ERROR_THREAD_QUEUE_DEADLOCK`. The error paths of both functions take
the lock a second time as well.

`rtems_rtl_elf_relocate_rel()` reads the addend of every relocation with an
aligned load before it knows the type. `R_MIPS_32` may target any byte offset,
so the processor takes an address error exception and the load of such an object
fails.

`dl06` fails wherever libdl is built, because `rtems-ld` and the RAP format have
no thread-local storage support. That is a property of the linker and not of a
BSP, and four BSPs state it individually while every other BSP reports a
failure. The init task of `dl14` needs the floating point attribute, which every
other test of that suite has.

Found while running the full test suite on eleven simulators. This description
was created with Claude Code assistance.

-- 
View it on GitLab: https://gitlab.rtems.org/rtems/rtos/rtems/-/work_items/5706
You're receiving this email because of your account on gitlab.rtems.org. 
Unsubscribe from this thread: 
https://gitlab.rtems.org/-/sent_notifications/5-5y8il61853av6b6lsyl5ykpls-1d/unsubscribe
 | Manage all notifications: https://gitlab.rtems.org/-/profile/notifications | 
Help: https://gitlab.rtems.org/help


_______________________________________________
bugs mailing list
[email protected]
http://lists.rtems.org/mailman/listinfo/bugs

Reply via email to