GNU bug report logs - #53878
[PATCH 00/11] Update Racket to 8.4. Adjust Chez Scheme

Previous Next

Full log

View this message in rfc822 format

From: Philip McGrath <philip.mcgrath <at> gmail.com>
To: Liliana Marie Prikler <liliana.prikler <at> gmail.com>, Philip McGrath <philip <at> philipmcgrath.com>, 53878 <at> debbugs.gnu.org, Ludovic Courtès <ludo <at> gnu.org>
Cc: Attila Lendvai <attila <at> lendvai.name>, Maxime Devos <maximedevos <at> telenet.be>, Malte Gerdes <malte.f.gerdes <at> gmail.com>, raingloom <raingloom <at> riseup.net>, zimoun <zimon.toutoune <at> gmail.com>
Subject: [bug#53878] [PATCH v3 09/15] gnu: Add racket-vm-cgc.
Date: Wed, 23 Feb 2022 13:55:16 -0500

Hi,

On 2/20/22 13:13, Liliana Marie Prikler wrote:
>> Given Ludo’'s explanation, I think the difference (or at least an
>> important difference) is that those functions are defined in `(guix
>> ...)` modules, as  opposed to `(gnu packages ...)` modules.
>>
>> But I wish I knew with any degree of certainty *why* this would be
>> true, if indeed it is.
>>
>> Maybe Maxime knows?
> Both Ludo and Maxime already explained this, but to be extra clear,
> it's the thunking.
> 
> (define foo <something-in-bar.scm>)
> (define bar <something-in-foo.scm>) <- problem
> 
> (define (foo) <something-in-bar.scm>)
> (define bar <something-in-foo.scm>) <- no problem.
> 
> Since inputs are thunked, you can define chez in chez.scm and racket in
> racket.scm and use them as input to each other without breaking the
> compiler (you will break the package builder though).  If you want
> something more meaningful, you can define racket-minimal in racket.scm
> and use it in chez.scm as input to a package, then use that package as
> input to racket.  This does not work for (source ) and (inherit )
> however, because those forms are not thunked.  You would have to thunk
> them until you reach a save haven (like a package's inputs), where you
> can call the thunk to produce a value.

To try to be concrete, I made the patch below as a mock-up of part of 
your earlier suggestion (IIUC):

On 2/20/22 04:03, Liliana Marie Prikler wrote:
> Inside chez-and-racket-bootstrap, define (make-<package>) functions for
> the following:
> - chez-bootstrap-bootfiles, chez-for-racket-bootstrap-bootfiles:
>    Taking version and origin.
> - racket-vm-cgc: Taking version and origin.
> - racket-vm-bc: Taking racket-vm-cgc.
> - racket-vm-cs: Taking racket-vm-bc.
>

...

>
> Inside racket, define %racket-version, %racket-origin, racket-minimal
> and racket.  It'd also be good if you made local definitions
> (define racket-vm-cgc (make-racket-vm-cgc %racket-version %racket-
> origin))
> (define racket-vm-bc (make-racket-vm-bc racket-vm-cgc))
> ...
> in this file.

This applies on top of v4, or I've put it at 
<https://gitlab.com/philip1/guix-patches/-/commit/982fe7cfb4d33103ee611acc310e3225ccf35852> 
if that's easier for anyone:

--8<---------------cut here---------------start------------->8---
From 982fe7cfb4d33103ee611acc310e3225ccf35852 Mon Sep 17 00:00:00 2001
From: Philip McGrath <philip <at> philipmcgrath.com>
Date: Wed, 23 Feb 2022 11:13:43 -0500
Subject: [PATCH] example of import problems

---
 gnu/packages/chez-and-racket-bootstrap.scm | 4 ++++
 gnu/packages/racket.scm                    | 3 +++
 2 files changed, 7 insertions(+)

diff --git a/gnu/packages/chez-and-racket-bootstrap.scm 
b/gnu/packages/chez-and-racket-bootstrap.scm
index b779099fb3..ea10f7fe92 100644
--- a/gnu/packages/chez-and-racket-bootstrap.scm
+++ b/gnu/packages/chez-and-racket-bootstrap.scm
@@ -47,6 +47,7 @@ (define-module (gnu packages chez-and-racket-bootstrap)
   #:use-module ((guix licenses)
                 #:prefix license:)
   #:export (chez-scheme-for-system
+            make-racket-vm-cgc
             racket-vm-for-system))

 ;; Commentary:
@@ -199,6 +200,9 @@ (define-module (gnu packages chez-and-racket-bootstrap)
 ;;
 ;; Code:

+(define (make-racket-vm-cgc a b)
+  42)
+
 (define* (chez-scheme-for-system #:optional
                                  (system (or (%current-target-system)
                                              (%current-system))))
diff --git a/gnu/packages/racket.scm b/gnu/packages/racket.scm
index c2854f84e8..08e437a722 100644
--- a/gnu/packages/racket.scm
+++ b/gnu/packages/racket.scm
@@ -58,6 +58,9 @@ (define %racket-version "8.4") ; MUST match 
"chez-and-racket-bootstrap.scm"
 (define %racket-commit
   (string-append "v" %racket-version))

+(define fake-racket-vm-cgc
+  (make-racket-vm-cgc 1 2))
+
 (define (extract-package-source origin spec)
   "Extract the source for a Racket package specified by SPEC from 
ORIGIN into
 a new file-like object.  In the resulting file-like object, the 
package source
-- 
2.32.0
--8<---------------cut here---------------end--------------->8---

With this patch, trying to run `make` fails with the same kind of error 
we've seen before:

```
ice-9/eval.scm:293:34: error: make-racket-vm-cgc: unbound variable
hint: Did you forget `(use-modules (gnu packages 
chez-and-racket-bootstrap))'?
```

Overall, I certainly agree that duplicating the definition of 
`%racket-version` is not ideal. I'd be glad for you or anyone to improve 
the situation, and I'll try to get my head around Maxime's email about 
the underlying semantics.

But I am confident that v4 of this series is at least not broken, if 
perhaps not maximally beautiful. Especially given that I, for one, have 
tried things that initially seemed correct only to discover subtle 
problems later, I think it would be better for any refinements to come 
in follow-on patches later.

-Philip

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.