Tier1 failure: test machine short of memory?

[Lindenmaier, Goetz]

Hi,

We could exclude it by setting
@requires !(sun.arch.data.model == "32")

The other test I mentioned, java/lang/String/StringRepeat.java, already
sets
@requires !(os.family == "windows" & sun.arch.data.model == "32")
but now fails on linux.

Best regards, Goetz

From: jdk-dev <jdk-dev-retn at openjdk.org> On Behalf Of Thomas Stüfe
Sent: Friday, March 15, 2024 2:20 PM
To: Severin Gehwolf <sgehwolf at redhat.com>
Cc: Andrew Haley <aph-open at littlepinkcloud.com>; jdk-dev at openjdk.org
Subject: Re: Tier1 failure: test machine short of memory?



On Fri, Mar 15, 2024 at 11:59 AM Severin Gehwolf <sgehwolf at redhat.com<mailto:sgehwolf at redhat.com>> wrote:
On Fri, 2024-03-15 at 10:12 +0000, Andrew Haley wrote:
> In java/util/HashMap/WhiteBoxResizeTest.jtr:
>
> Error occurred during initialization of VM
> Could not reserve enough space for 2097152KB object heap
>
> So we're seeing red on the tier1 runs on Linux-i686. I'm not sure what we can
> do about this, perhaps we should disable the test for x86-32?

The test seems to need a 2GB heap:

run=USER_SPECIFIED testng/othervm -Xmx2g WhiteBoxResizeTest

So perhaps this as a fix?

@requires (os.maxMemory > 2g)

A 2GB allocation on 23-bit is more likely to run out of address space than out of physical memory.

I think excluding this for 32-bit makes the most sense.


Thanks,
Severin
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.org/pipermail/jdk-dev/attachments/20240315/740ed0e2/attachment.htm>