RFR(S) 8231986 [SA] Consolidate parts of the Linux and MacOSX versions of ps_core.c

[Ioi Lam]

Hi Chris,

Thanks for the review. I moved the declaration of n and m to inside the 
"if" block. For ps_pread vs ps_pdread, I am not sure why the symbol is 
different on mac vs linux. Maybe the purpose is to avoid conflicting 
with built-in symbols on the platform? Since this function is used by 
other parts of SA, I think I'll just leave it as is.

Thanks
- Ioi


On 10/9/19 11:39 AM, Chris Plummer wrote:
> Hi Ioi,
>
> Overall the changes look fine. A couple of minor things:
>
> Any reason not to fix the ps_pread/ps_pdread naming issue rather than 
> just map it with a #define?
>
> In init_classsharing_workaround(), I think the "n" and "m" 
> declarations should be inside the "if" block.
>
> thanks,
>
> Chris
>
> On 10/7/19 11:37 PM, Ioi Lam wrote:
>> https://bugs.openjdk.java.net/browse/JDK-8231986
>> http://cr.openjdk.java.net/~iklam/jdk14/8231986-consolidate-ps-core.v01/
>>
>> One of my upcoming CDS changes (JDK-8231610) would affect the 
>> duplicated code
>> in these 2 files. So instead of fixing the same thing twice, I have 
>> moved the
>> duplicated parts of these files that are related to CDS into a common 
>> file.
>> This would simplify future maintenance of CDS+SA code.
>>
>> To make my life simple, I just moved all functions up to 
>> init_classsharing_workaround().
>> The remaining lines of the these 2 files seem to be unrelated to 
>> hotspot (core file
>> handling, elf, etc), so they probably don't need to be changed any 
>> time soon. I'll leave
>> the duplications there as is.
>>
>> Thanks
>> - Ioi
>