RFR: Store AOT code in cached code region of AOT cache [v3]

[Vladimir Kozlov]

On Wed, 12 Feb 2025 00:07:43 GMT, Ashutosh Mehra <asmehra at openjdk.org> wrote:

>> This PR stores AOT code in the cached code region of AOT Cache.
>> 
>> Some things worth pointing out:
>> 1. Before this patch: the AOT code is first generated in a temporary buffer. After all the code is generated, this temporary buffer is copied into second buffer (allocated in native heap) for multiple reasons. This second buffer is then stored in the file as is. With this patch the second buffer is now allocated in the cached code region which is part of the AOT cache. So now the cached code region contains the permanent oops table (which already exists before this patch) and the AOT code+data.
>> 2. I have removed `_jvm_version_offset` from the SCCHeader. AOT Cache already stores VM identifier which should cover it.
>> ~3. I have to remove the code that maps cached code region to CodeCache because the contents in cached code region are not CodeBlobs. The cached code region can be mapped to any random address.~
>> 4. I have replaced references to "Startup Code Cache" with "AOT Code Cache" or "AOT Code".
>
> Ashutosh Mehra has updated the pull request incrementally with one additional commit since the last revision:
> 
>   Map cached code region separately from the CodeCache
>   
>   Signed-off-by: Ashutosh Mehra <asmehra at redhat.com>

src/hotspot/share/cds/archiveBuilder.cpp line 315:

> 313: 
> 314: address ArchiveBuilder::reserve_buffer() {
> 315:   size_t buffer_size = LP64_ONLY(CompressedClassSpaceSize) NOT_LP64(256 * M) + ReservedCodeCacheSize;

Please add comment about `ReservedCodeCacheSize`. Why not `CachedCodeMaxSize`?

src/hotspot/share/code/codeCache.cpp line 359:

> 357: }
> 358: 
> 359: 

You can undo that to avoid changing file.

test/hotspot/jtreg/premain/lib/DemoSupport.gmk line 163:

> 161: 	     ${DEMO_CMDLINE}
> 162: 	ls -l ${DEMO_CDS}
> 163: 	#ls -l ${DEMO_CDS}.code

I would just remove this line.

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

PR Review Comment: https://git.openjdk.org/leyden/pull/39#discussion_r1952897696
PR Review Comment: https://git.openjdk.org/leyden/pull/39#discussion_r1952904753
PR Review Comment: https://git.openjdk.org/leyden/pull/39#discussion_r1952905831