[Adeos-main] [pull request] Avoid double bug reports

[Adeos-main] [pull request] Avoid double bug reports

[Jan Kiszka]

The following changes since commit 88fd7426f3565647b4f5ef93d8e612e059dae873:
  Philippe Gerum (1):
        Make generic atomic ops Adeos-aware

are available in the git repository at:

  git://git.kiszka.org/ipipe-2.6.git queues/2.6.31-noarch

Jan Kiszka (2):
      noarch: Don't check for internal errors during oops
      noarch: Set oops_in_progress on failing context check

 kernel/ipipe/core.c |    6 +++---
 1 files changed, 3 insertions(+), 3 deletions(-)

-------

noarch: Don't check for internal errors during oops

We are busted already, verifying the context here only generates
confusing output or causes even more corruptions.

Signed-off-by: Jan Kiszka <jan.kiszka@domain.hid>
---
 kernel/ipipe/core.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/kernel/ipipe/core.c b/kernel/ipipe/core.c
index ffaceaa..c9bc5b4 100644
--- a/kernel/ipipe/core.c
+++ b/kernel/ipipe/core.c
@@ -312,7 +312,7 @@ void __ipipe_unstall_root(void)
 
 #ifdef CONFIG_IPIPE_DEBUG_INTERNAL
 	/* This helps catching bad usage from assembly call sites. */
-	BUG_ON(!__ipipe_root_domain_p);
+	BUG_ON(!__ipipe_root_domain_p && !oops_in_progress);
 #endif
 
 	p = ipipe_root_cpudom_ptr();
@@ -328,7 +328,7 @@ void __ipipe_unstall_root(void)
 void __ipipe_restore_root(unsigned long x)
 {
 #ifdef CONFIG_IPIPE_DEBUG_INTERNAL
-	BUG_ON(!ipipe_root_domain_p);
+	BUG_ON(!ipipe_root_domain_p && !oops_in_progress);
 #endif
 
 	if (x)
-- 
1.6.0.2

------

noarch: Set oops_in_progress on failing context check

This helps to avoid raising multiple bugs that may even precede the
actually issue in the output.

Signed-off-by: Jan Kiszka <jan.kiszka@domain.hid>
---
 kernel/ipipe/core.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/kernel/ipipe/core.c b/kernel/ipipe/core.c
index c9bc5b4..aa2e5a0 100644
--- a/kernel/ipipe/core.c
+++ b/kernel/ipipe/core.c
@@ -1676,7 +1676,7 @@ void ipipe_check_context(struct ipipe_domain *border_domain)
 
 	ipipe_context_check_off();
 	ipipe_trace_panic_freeze();
-	ipipe_set_printk_sync(__ipipe_current_domain);
+	oops_in_progress = 1;
 
 	if (this_domain->priority > border_domain->priority)
 		printk(KERN_ERR "I-pipe: Detected illicit call from domain "
-- 
1.6.0.2

Re: [Adeos-main] [pull request] Avoid double bug reports

[Gilles Chanteperdrix]

Jan Kiszka wrote:
> The following changes since commit 88fd7426f3565647b4f5ef93d8e612e059dae873:
>   Philippe Gerum (1):
>         Make generic atomic ops Adeos-aware
> 
> are available in the git repository at:
> 
>   git://git.kiszka.org/ipipe-2.6.git queues/2.6.31-noarch
> 
> Jan Kiszka (2):
>       noarch: Don't check for internal errors during oops
>       noarch: Set oops_in_progress on failing context check
> 
>  kernel/ipipe/core.c |    6 +++---
>  1 files changed, 3 insertions(+), 3 deletions(-)
> 
> -------
> 
> noarch: Don't check for internal errors during oops
> 
> We are busted already, verifying the context here only generates
> confusing output or causes even more corruptions.
> 
> Signed-off-by: Jan Kiszka <jan.kiszka@domain.hid>

Are we sure this fixes all issues? Maybe we should wait for Josh tests
before releasing another I-pipe patch?

-- 
                                          Gilles

Re: [Adeos-main] [pull request] Avoid double bug reports

[Jan Kiszka]

Gilles Chanteperdrix wrote:
> Jan Kiszka wrote:
>> The following changes since commit 88fd7426f3565647b4f5ef93d8e612e059dae873:
>>   Philippe Gerum (1):
>>         Make generic atomic ops Adeos-aware
>>
>> are available in the git repository at:
>>
>>   git://git.kiszka.org/ipipe-2.6.git queues/2.6.31-noarch
>>
>> Jan Kiszka (2):
>>       noarch: Don't check for internal errors during oops
>>       noarch: Set oops_in_progress on failing context check
>>
>>  kernel/ipipe/core.c |    6 +++---
>>  1 files changed, 3 insertions(+), 3 deletions(-)
>>
>> -------
>>
>> noarch: Don't check for internal errors during oops
>>
>> We are busted already, verifying the context here only generates
>> confusing output or causes even more corruptions.
>>
>> Signed-off-by: Jan Kiszka <jan.kiszka@domain.hid>
> 
> Are we sure this fixes all issues? Maybe we should wait for Josh tests
> before releasing another I-pipe patch?

I doesn't fix Josh's primary issue, it targets at some of the succeeding
errors he got (it doesn't make sense to complain about incorrect
contexts twice or more, specifically if the BUG_ONs provide less
information than the primary check).

Jan

-- 
Siemens AG, Corporate Technology, CT T DE IT 1
Corporate Competence Center Embedded Linux