harmony-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Evgueni Brevnov" <evgueni.brev...@gmail.com>
Subject Re: [drlvm] jvmti tests crash on suse9 ia32
Date Wed, 29 Nov 2006 05:44:05 GMT
On 11/28/06, Gregory Shimansky <gshimansky@gmail.com> wrote:
> Geir Magnusson Jr. wrote:
> > Sorry?   What were the recent changes?
>
> Evgueni probably mean the patch in HARMONY-2320 which I've committed
> yesterday. It is actually a change to *drlvm* Windows *exceptions*
> handling. As far as I understand Linux signals are not touched yet.

Sorry for not being clear. I was talking about changes Jeff made in
classlib's code. See his message below. I read the message once again
and realized that the fix was not propagated to the Harmony code
:-(....So it can't be the reason of the crash of jvmti tests. Anyway I
wish to see this fix in Harmony....:-)

Evgueni

Jeff Disher wrote:
We fixed mapPortLibSignalToUnix in our code but it required changing the
interface to return a signed integer so that we could return an "unmatched"
value which immediately would be obviously recognized as being outside the
range of valid return values.

The key point here is that it was a bug and there was no special logic to
handle any of those other signal numbers outside of this mechanism.  We
hadn't historically used them so there was no mapping defined.

This bug should be fixed in Harmony, as well,
Jeff.

>
> > Evgueni Brevnov wrote:
> >> It could be a result of recent changes in classlib's signal handling
> >> code...
> >>
> >> Evgueni
> >>
> >> On 11/28/06, Alexei Fedotov <alexei.fedotov@gmail.com> wrote:
> >>> Alexey,
> >>> I cannot find a JIRA which mentions masterSynchSignalHandler, which to
> >>> my mind differentiates the stack trace. Probably it worth to file a
> >>> new one.
> >>>
> >>> --
> >>> Thank you,
> >>> Alexei
> >>>
> >>> On 11/27/06, Alexey Varlamov <alexey.v.varlamov@gmail.com> wrote:
> >>> > Is this a known regression?
> >>> >
> >>> >    [junit] An unhandled error (64) has occurred.
> >>> >    [junit] HyGeneric_Signal_Number=00000040
> >>> >    [junit] Signal_Number=00000005
> >>> >    [junit] Error_Value=00000000
> >>> >    [junit] Signal_Code=00000080
> >>> >    [junit] Handler1=0804A7E8
> >>> >    [junit] Handler2=40026667
> >>> >    [junit] EDI=00000056
> >>> >    [junit] ESI=00000001
> >>> >    [junit] EAX=00000009
> >>> >    [junit] EBX=40CD4444
> >>> >    [junit] ECX=00000009
> >>> >    [junit] EDX=00000005
> >>> >    [junit] EIP=5651AE70
> >>> >    [junit] ES=0000007B
> >>> >    [junit] DS=0000007B
> >>> >    [junit] ESP=BFFFC96C
> >>> >    [junit] EFlags=00000206
> >>> >    [junit] CS=00000073
> >>> >    [junit] SS=0000007B
> >>> >    [junit] EBP=BFFFCA84
> >>> >    [junit] SIGABRT in VM code.
> >>> >    [junit] Stack trace:
> >>> >    [junit]     1: ?? (??:-1)
> >>> >    [junit]     2: abort (??:-1)
> >>> >    [junit]     3: masterSynchSignalHandler (??:-1)
> >>> >    [junit] <end of stack trace>
> >>> >    [junit] Test Breakpoint1.Breakpoint1 FAILED
> >>> >
> >>> > BUILD FAILED
> >>> >
> >>>
> >
>
>
> --
> Gregory
>
>

Mime
View raw message