Module: xenomai-jki
Branch: for-upstream
Commit: 0b422dc91ee78c1d2dca74280ce857a9c4c14748
URL:    
http://git.xenomai.org/?p=xenomai-jki.git;a=commit;h=0b422dc91ee78c1d2dca74280ce857a9c4c14748

Author: Jan Kiszka <jan.kis...@siemens.com>
Date:   Sun Sep 11 12:07:03 2011 +0200

nucleus: Remove user space triggerable bug check

The rescnt maintenance depends on the ABI convention that SCHED_OTHER
user space threads enter the kernel for resource allocation and release
unconditionally. The current logic also does not account for threads
switching from SCHED_FF/RR to SCHED_OTHER while holding locks.

So triggering a BUG on this counter becoming < 0 is a bad idea, even if
we fixed the policy switch issue.

Signed-off-by: Jan Kiszka <jan.kis...@siemens.com>

---

 ksrc/nucleus/synch.c |    1 -
 1 files changed, 0 insertions(+), 1 deletions(-)

diff --git a/ksrc/nucleus/synch.c b/ksrc/nucleus/synch.c
index b956e46..93ba988 100644
--- a/ksrc/nucleus/synch.c
+++ b/ksrc/nucleus/synch.c
@@ -686,7 +686,6 @@ xnsynch_release_thread(struct xnsynch *synch, struct 
xnthread *lastowner)
 
        if (xnthread_test_state(lastowner, XNOTHER))
                xnthread_dec_rescnt(lastowner);
-       XENO_BUGON(NUCLEUS, xnthread_get_rescnt(lastowner) < 0);
        lastownerh = xnthread_handle(lastowner);
 
        if (use_fastlock &&


_______________________________________________
Xenomai-git mailing list
Xenomai-git@gna.org
https://mail.gna.org/listinfo/xenomai-git

Reply via email to