[EXTERNAL] Re: Proposed JEP - Deprecate the Windows x86-32 Port

[Alex Buckley]

Hi George,

A couple of stylistic points on the JEP:

1. Goals and Non-Goals usually appear before Motivation, e.g., as done 
in JEP 362.

2. In Alternatives, there is a paragraph about "It is also known that 
32-bit JVMs on Windows ...". This paragraph describes what a user could 
do if this JEP proceeds, but Alternatives is usually about what the JEP 
author could do instead the thing they just spelled out in the Description.

I recommend moving the paragraph to Risks & Assumptions, where you 
already have a user story about WOW64. You could say something like the 
text below. I made numerous edits for tone (I thought "in the wild" was 
a bit flippant) and for clarity (there was an "older" and "newer" theme 
going on that I thought unnecessary) and most importantly to spell out 
the assumption.

-----
It is known that 32-bit JVMs on Windows are still used due to 
integration with 32-bit DLLs. These users cannot migrate directly to 
64-bit JVMs, because a 64-bit process on Windows cannot load a 32-bit 
DLL. This JEP assumes that these users can continue to run a 32-bit JVM 
to integrate with these native libraries, and expose their functionality 
over some form of remote API that can be consumed by code in a 64-bit 
JVM, within the same environment.
-----

Alex

On 3/3/2023 2:43 AM, George Adams wrote:
> Thank you so much for the feedback that everyone has provided so far. I 
> have pushed a handful of changes to the JEP draft to help clarify some 
> points. Most notably, this JEP does not intend to deprecate any other 
> x86_32 ports.
> 
> Thanks
> 
> George Adams
> 
> *From: *Glavo <zjx001202 at gmail.com>
> *Date: *Wednesday, 1 March 2023 at 12:08
> *To: *Alan Bateman <Alan.Bateman at oracle.com>
> *Cc: *George Adams <George.Adams at microsoft.com>, 
> jdk-dev at openjdk.java.net <jdk-dev at openjdk.java.net>
> *Subject: *[EXTERNAL] Re: Proposed JEP - Deprecate the Windows x86-32 Port
> 
> 	
> 
> You don't often get email from zjx001202 at gmail.com. Learn why this is 
> important <https://aka.ms/LearnAboutSenderIdentification>
> 
> 	
> 
> I have seen many Intel Celeron N/J machines with memory <4GiB. Although 
> x86-32 is old, its non-heap memory footprint is lower, so it works 
> better under resource constraints.
> 
> And although Windows 11 does not provide 32-bit images, as far as I 
> know, Microsoft does not plan to give up support for running 32-bit 
> programs on 64-bit systems.
> 
> 32-bit programs are still an important part of Windows, a large number 
> of programs are still 32-bit.
> 
> For a considerable number of client applications, upgrading to 64-bit 
> has only negative benefits.
> 
> They only need a little memory and are not sensitive to performance.
> 
> What will 64-bit bring to them? Higher resource consumption and poor 
> compatibility (cannot run on x86-32 or Windows 10 on Arm systems)
> 
> I know that maintaining a port requires a lot of manpower, so I can't 
> ask you to do anything. But I really hope it will continue to be maintained.
> 
> On Tue, Feb 28, 2023 at 3:15 AM Alan Bateman <Alan.Bateman at oracle.com 
> <mailto:Alan.Bateman at oracle.com>> wrote:
> 
>     On 27/02/2023 11:04, George Adams wrote:
> 
>         Hi all,
> 
>         I’ve been asked to socialize my proposed JEP to deprecate the
>         Windows x86-32 port on this mailing list.
> 
>         A link to the draft JEP can be found here:
>         https://bugs.openjdk.org/browse/JDK-8303167
>         <https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fbugs.openjdk.org%2Fbrowse%2FJDK-8303167&data=05%7C01%7CGeorge.Adams%40microsoft.com%7C645a895d36614b8ebfa408db1a4db724%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C638132693333901376%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=DdaeznIQYUT4W4SQ4tlMDVUs3pQupv5mG2LrUYXN6J0%3D&reserved=0>
> 
>         In summary, the main motivation for this JEP is that there is
>         currently no implementation of JEP 436 (Virtual Threads)
>         <https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fopenjdk.org%2Fjeps%2F436&data=05%7C01%7CGeorge.Adams%40microsoft.com%7C645a895d36614b8ebfa408db1a4db724%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C638132693333901376%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000%7C%7C%7C&sdata=QD41i%2B7kj3j01c5%2B3b0pRSx6NKY751CUnfrele7Har0%3D&reserved=0> for 32-bit platforms and without a vendor stepping forward to implement this it's unlikely that OpenJDK will be able to continue supporting 32-bit architectures. Another motivation is that Windows 10 (the last Windows operating system to support a 32-bit installation) will reach EOL on October 14, 20251.
> 
>     When you build JDK 19+ to target windows-x86 then it will use an
>     alternative implementation of virtual thread that creates a kernel
>     thread for each virtual thread. So it doesn't scale but it's good
>     enough for Zero and ports that are a bit behind.
> 
>     That said, it's a good topic to bring up. I don't expect dropping
>     windows-x86 will remove the burden of keeping the x86_32 port
>     working, to do that would require dropping linux-x86 too. So maybe
>     the discussion should be broadened to ask if the time is approaching
>     to remove the x86_32 port? At one point, one of the arguments to
>     keep linux-x86 working was reconditioning older computers but I
>     don't know if this is still the case. I see a mail to jdk-dev from
>     Mark Yagnatinsky that talks about JNI libs or drivers that are
>     32-bit only. There isn't much context but it would be surprising for
>     something that is actively maintained to not have a 64-bit build in
>     2023. He also mentions limiting resources but that may be a case
>     where an OS container should be used. It might be that you expand
>     the Motivation in draft JEP to cover these points.
> 
>     -Alan
>