Request for review (S): 7091417 recvfrom's 6th input should be of type socklen_t

[Paul Hohensee]

Then that's a short term workaround.

Also, if your proposed change is the minimum you can get away with to
build on s11, then you could use the CR# you've filed to fix it: it won't
make things worse.  If you want to do that, consider it reviewed.

Paul

On 10/19/11 11:14 AM, Christian Thalinger wrote:
> Yes, indeed.  -- Chris
>
> On Oct 19, 2011, at 4:44 PM, Paul Hohensee wrote:
>
>> Can you build on S10?
>>
>> On 10/19/11 10:32 AM, Christian Thalinger wrote:
>>> Since I forgot you are already working on this and sent another bug report today, were are we with that fix?
>>>
>>> -- Chris
>>>
>>> On Sep 23, 2011, at 11:48 PM, Paul Hohensee wrote:
>>>
>>>> And, jvm.h is an exported-to-the-jdk file, so there are probably build issues
>>>> there as well.  Not to mention that "accept" and some other socket operations
>>>> want socklen_t* too.  I'm working with Azeem on a more extensive fix.
>>>>
>>>> Paul
>>>>
>>>> On 9/23/11 5:29 PM, John Coomes wrote:
>>>>> Azeem Jiva (azeem.jiva at oracle.com) wrote:
>>>>>> 7091417:  recfrom's 6th input should be socklen_t
>>>>>>
>>>>>> The current recvfrom implementation won't compile on Solaris 11 since the 6th
>>>>>> input should be of type socklen_t.  Diffs are attached
>>>>>> ...
>>>>> Did you build on windows?  Don't think socklen_t exists there.  Also
>>>>> the os.hpp decl of recvfrom() doesn't match the definition in
>>>>> os_windows.cpp anymore.
>>>>>
>>>>> -John