RFR: 8319873: Add windows implementation for jcmd System.map and System.dump_map [v13]

Simon Tooke stooke at openjdk.org
Tue Sep 17 14:28:11 UTC 2024


On Tue, 17 Sep 2024 13:17:49 GMT, Kevin Walls <kevinw at openjdk.org> wrote:

>> Simon Tooke has updated the pull request incrementally with one additional commit since the last revision:
>> 
>>   remove trailing whitespace
>
> Thanks Thomas - 
> It's understood what the info is, main question was when is it useful, and do we need to differ from Linux.
> 
> Looking at 8318636: https://github.com/openjdk/jdk/pull/16301
> the Linux example linked in the introduction there shows an offset column.
> 
> Maybe Simon you're taking inspiration from that original, 
> and there was an extension to the feature just over a month ago in
> 8322475: https://github.com/openjdk/jdk/pull/17158
> 
> ...where the offset column was taken out.
> 8322475 also separated "info" and "file" into their own columns.  Only CDS and one of the hsperfdata lines use both columns that I notice, but there may be others.
> Here in Windows we're going with one column for both, unlike the updated Linux format.
> 
> This PR is implementing "classic" System.map, not the "modern 2024" version? 8-)
> 
> It's a good feature to have, I am just being annoyingly clear about when and why we differ between platforms.

@kevinjwalls I believe we all expect that information will be different between platforms, and some information is more important on one platform than another.  For example I have a rough implementation of a macOS System.map and it will differ again, although as the same author, it will have some similarities (e.g. merged vminfo/file field) to this implementation.  I agree the offset column isn't vital, but the start of allocation may be (at some point, to someone), and, as @tstuefe commented, there is no equivalent on Linux.

-------------

PR Comment: https://git.openjdk.org/jdk/pull/20597#issuecomment-2356011117


More information about the serviceability-dev mailing list