From jahoda at aicas.de  Mon Oct  1 12:44:28 2018
From: jahoda at aicas.de (Ladislav Jahoda)
Date: Mon, 1 Oct 2018 14:44:28 +0200
Subject: Infinite loop in jdk/test/sun/management/windows/revokeall
Message-ID: <AD381EDF44CB4BE0B8D7D9BA9E931F3A@ladaPC>

Hello guys,

the revokeall utility has an infinite loop in its revokeAll method. Please note in the loop (line #344 in revokeall.c):

while (count > 0) {

the condition (line #354):

if (((ACCESS_ALLOWED_ACE *)ace)->Header.AceType != ACCESS_ALLOWED_ACE_TYPE) {
     continue;
}

which loops forever if the condition is met. Note that the ?count? counter needs to be decremented before the continue statement, that is:

if (((ACCESS_ALLOWED_ACE *)ace)->Header.AceType != ACCESS_ALLOWED_ACE_TYPE) {
     count--;
     continue;
}

I checked the OpenJDK 8, 9 and 10, the bug is in all versions.

Thanks ahead for the fix,
Lada Jahoda


 
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/jtreg-use/attachments/20181001/6dc061be/attachment.html>

From jonathan.gibbons at oracle.com  Mon Oct  1 14:56:56 2018
From: jonathan.gibbons at oracle.com (Jonathan Gibbons)
Date: Mon, 1 Oct 2018 07:56:56 -0700
Subject: Infinite loop in jdk/test/sun/management/windows/revokeall
In-Reply-To: <AD381EDF44CB4BE0B8D7D9BA9E931F3A@ladaPC>
References: <AD381EDF44CB4BE0B8D7D9BA9E931F3A@ladaPC>
Message-ID: <5b29307d-903d-190d-c074-7be5955910ea@oracle.com>

I'm guessing serviceability-dev is a better place for this message.

-- Jon



On 10/1/18 5:44 AM, Ladislav Jahoda wrote:
> Hello guys,
> the revokeall utility has an infinite loop in its revokeAll method. 
> Please note in the loop (line #344 in revokeall.c):
> while (count > 0) {
> the condition (line #354):
> if (((ACCESS_ALLOWED_ACE *)ace)->Header.AceType != 
> ACCESS_ALLOWED_ACE_TYPE) {
> ???? continue;
> }
> which loops forever if the condition is met. Note that the ?count? 
> counter needs to be decremented before the continue statement, that is:
> if (((ACCESS_ALLOWED_ACE *)ace)->Header.AceType != 
> ACCESS_ALLOWED_ACE_TYPE) {
> ???? count--;
> ???? continue;
> }
> I checked the OpenJDK 8, 9 and 10, the bug is in all versions.
> Thanks ahead for the fix,
> Lada Jahoda
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.openjdk.java.net/pipermail/jtreg-use/attachments/20181001/e20aee62/attachment.html>

From gerard.ziemski at oracle.com  Wed Oct 17 21:54:20 2018
From: gerard.ziemski at oracle.com (Gerard Ziemski)
Date: Wed, 17 Oct 2018 21:54:20 -0000
Subject: Does jtreg ever send SIGSTOP to a test?
Message-ID: <3629CA30-BE88-4603-845C-76C4872AC035@oracle.com>

hi guys,

I?m investigating https://bugs.openjdk.java.net/browse/JDK-8210184 ?runtime/4243535/simple.sh failed with Segmentation fault: 11? and according to the corefile, the test received a SIGSTOP signal.

Does jtreg signal tests? Would it ever use SIGSTOP to stop a test, and if so why?

Any idea where such signal could have came from?


cheers