RFR(XS): 8008255: jvmtiExport.cpp::post_to_env() does not check malloc() return

[Daniel D. Daugherty]

Reduced the audience to just the three of us...

Rickard, code review requests are always supposed to have at least a
one line blurb about what testing was done. For this bug, I would have
said something like this:

     This bug was found by code inspection so there is no existing test
     associated with this bug. It is very difficult to create an automated
     test for a memory allocation failure. Standard JPRT testing was done.

Something like the above (or something shorter) would probably have
forestalled any questions about testing during the review.

Ron, your previous work on the memory allocation code required more
testing because you were changing the mechanism. In particular, you
were changing the mechanism as it related to multi-threaded allocation
failure. That change definitely required the temporary testing hook and
manual testing that you did. In this bug fix, Rickard is just using the
vm_exit_out_of_memory() call. He isn't changing it.

Dan


On 5/8/13 6:33 AM, Rickard Bäckman wrote:
> Ron,
>
> Thanks for the review.
>
> I ran the change through JPRT.
> Creating a specific test for this seems rather problematic as we need to make sure that a specific malloc() fails.
>
> /R
>
> On May 8, 2013, at 2:31 PM, Ron Durbin wrote:
>
>> Change looks good, can tell me what tests were run?
>> Were able create a unit test that triggers the failure ?
>>
>> Thx Ron
>>
>>> -----Original Message-----
>>> From: Staffan Larsen
>>> Sent: Wednesday, May 08, 2013 6:17 AM
>>> To: Rickard Bäckman
>>> Cc: serviceability-dev at openjdk.java.net serviceability-dev at openjdk.java.net; hotspot-runtime-
>>> dev at openjdk.java.net
>>> Subject: Re: RFR(XS): 8008255: jvmtiExport.cpp::post_to_env() does not check malloc() return
>>>
>>> Looks good.
>>>
>>> /Staffan
>>>
>>> On 8 maj 2013, at 13:08, Rickard Bäckman <rickard.backman at oracle.com> wrote:
>>>
>>>> Hi,
>>>>
>>>> please review this small change.
>>>> The change adds a check to see the result of a malloc() and aborts the VM with an out of memory
>>> error.
>>>> Webrev: http://cr.openjdk.java.net/~rbackman/8008255/
>>>> Bug: http://bugs.sun.com/view_bug.do?bug_id=8008255
>>>>
>>>> Thanks
>>>> /R
>>>>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://mail.openjdk.java.net/pipermail/hotspot-runtime-dev/attachments/20130508/d489937a/attachment.html