Chez Scheme for Racket build on aarch64 (patch attached)

Hi Guix Team,

tl;dr;

The recipe for package 'chez-scheme-for-racket-bootstrap-bootfiles' (a

dependency for racket-8.*) needs to use the --machine flag to build on aarch64

(and possibly also ppc32, the other non-x86 architectures supported by the

Racket fork of Chez Scheme). I therefore suggest the attached patch to make

building and using Racket possible on aarch64 Guix.

End of tl;dr;

A few weeks ago I wanted to use Racket on Guix on aarch64, but was made aware,

that Racket was not available as a substitute and it did not build. On the

other hand, recent versions of Racket were available on Raspberry Pi OS which is

explicitly running on aarch64. The machine I am running Guix on actually is a

Raspberry Pi 4B.

So, I dug deeper to find that the problem is that the package

chez-scheme-for-racket-bootstrap-bootfiles

fails with the error

,----

| string-append: contract violation

| expected: string?

| given: #f

| argument position: 1st

| other arguments...:

| ".def"

| context...:

| /tmp/guix-build-chez-scheme-for-racket-bootstrap-bootfiles-9.9.9-pre-release.14.drv-0/source/racket/src/rktboot/machine-def.rkt:6:0: open-file-with-machine.def-redirect

| error: in phase 'build': uncaught exception:

| %exception #<&invoke-error program: "/gnu/store/0n4skakcg05bbq9sq7g00j2zm25xh4wa-racket-vm-bc-8.8/opt/racket-vm/bin/racket" arguments: ("../rktboot/main.rkt") exit-status: 1 term-signal: #f stop-signal: #f>

| phase `build' failed after 14.3 seconds

| command "/gnu/store/0n4skakcg05bbq9sq7g00j2zm25xh4wa-racket-vm-bc-8.8/opt/racket-vm/bin/racket" "../rktboot/main.rkt" failed with status 1

| builder for `/gnu/store/d04mbj02vzjv0ljcqy2lxs33scz7kjz0-chez-scheme-for-racket-bootstrap-bootfiles-9.9.9-pre-release.14.drv' failed with exit code 1

| @ build-failed /gnu/store/d04mbj02vzjv0ljcqy2lxs33scz7kjz0-chez-scheme-for-racket-bootstrap-bootfiles-9.9.9-pre-release.14.drv - 1 builder for `/gnu/store/d04mbj02vzjv0ljcqy2lxs33scz7kjz0-chez-scheme-for-racket-bootstrap-bootfiles-9.9.9-pre-release.14.drv' failed with exit code 1

`----

(see, e.g. the build [442314 on ci.guix.gnu.org]).

The Racket fork of Chez is natively supported on aarch64, which is supported by

comments in module (gnu packages chez), so I went on to find that machine

architecture is communicated in the build process by setting the environment

variable MACH. I haven't looked to verify this, but it seems that this

environment variable is normally set to a guessed value by the zuo build process

(zuo for 'make' in Mandarin?).

The Guix build recipe for chez-scheme-for-racket-bootstrap-bootfiles, though, is

not configuring the whole downloaded racket source package but takes the

shortcut of just building ChezScheme in the subdirectory

racket/src/ChezScheme

using the Racket utility

../rktboot/main.rkt

This script would actually accept the flag '--machine' and would set the

environment variable MACH to the value of the argument following the flag.

The problem is that in the case of aarch64 the recipe would not set the flag.

In the recipe for chez-scheme-for-racket-bootstrap-bootfiles in module (gnu

packages chez), we find the lines

,----

| #~(invoke

| (search-input-file (or native-inputs inputs)

| "/opt/racket-vm/bin/racket")

| "../rktboot/main.rkt"

| #$@(if (racket-cs-native-supported-system?)

| #~()

| (let ((m (nix-system->pbarch-machine-type)))

| #~("--machine" #$m)))))))))))))

`----

and, since `(racket-cs-native-supported-system?)' returns `tarm64le', the first

branch of the if expression is taken, ommitting the flag.

I suggest to change the above code to the following:

,----

| #~(invoke

| (search-input-file (or native-inputs inputs)

| "/opt/racket-vm/bin/racket")

| "../rktboot/main.rkt"

| #$@(let ((m (or (racket-cs-native-supported-system?)

| (nix-system->pbarch-machine-type))))

| #~("--machine" #$m))))))))))))

`----

which sets the flag for natively supported architectures and those supported by

portable bytecode.

The package (and subsequently Racket) builds on aarch64, the build on x86_64

seems to be unaffected.

It is a simple change which would make a huge difference for Guix users who want

to be able to play with Racket on an aarch64 machine.

Cheers

Tim

Lá Fhéile Pádraig sona dhaoibh!

[442314 on ci.guix.gnu.org] https://ci.guix.gnu.org/build/442314/log/raw

Hi Tim,

On Thursday, March 16, 2023 7:23:46 PM EDT Tim Johann wrote:

Thanks again for your investigation of this problem! As I commented on the

upstream issue at https://github.com/racket/racket/issues/3948, I still

think this is an upstream bug (and your analysis helped a great deal in finding

the root cause), but the approach in your patch is probably the easiest way to

fix things for Guix users for now. I'd just ask that you add a comment

referencing the upstream issue and with some explanation, because it's not my

preferred long-term approach.

The "rktboot" scripts (which simulate enough of Chez Scheme to bootstrap its

bootfiles) are supposed to be able to infer right machine type for the current

platform, at least if the platform is natively supported. The problem is that

the upstream code to doesn't actually work on architectures other than i386

and x86_64.

This is a good work-around. The reason I'd like to avoid it in the long run is

that, ideally, I'd like to get us out of the business of managing Chez machine

types and the translation from normal architecture and OS strings. I think

we'll at least be able to eliminate `%chez-features-table` once the ongoing

work to reunify upstream Chez Scheme with Racket's branch is complete.

Reducing the number of places where we have to use these machine-type-

producing functions should help advance that goal, so I'd prefer to have a

comment here reminding us that we can reconsider "--machine" once https://

github.com/racket/racket/issues/3948> is fixed.

-Philip

Hi Philip,

Happy to be of help.

Thanks for the pointers, here's the amended patch. I hope it is to your liking.

cheers

Tim

On Friday, March 24, 2023 4:03:19 PM EDT Tim Johann wrote:

Thanks, this addresses my concerns!

I'm not a Guix committer, so I'm cc'ing a couple of people who have applied

recent patches to this file. They may want to review the commit message for

style, also.

I expect the underlying issue would also affect riscv64, but I hit some network

trouble when trying to look for a build log just now.

-Philip

Hi,

Tim Johann <t1m@phrogstar.de> skribis:

Applied!

Thank you Tim, and thanks Philip for reviewing.

Tim: please use ‘git format-patch’ in the future; the format here cannot

be ingested as-is by ‘git am’.

Ludo’.