gnucash-1.8.8 crashes on start-up with slib-3.1.1 reverting to slib-2.4.6 fixes the problem. I think a function in slib-2.4.6 is missing in slib-3.1.1 running gnucash at the command line yields: gnucash Backtrace: In unknown file: ... ?: 37 (begin (if # #) (make-modules-in # full-name)) ?: 38* (if (or # #) (try-load-module name)) ?: 39 [try-load-module (gnucash report iframe-url)] ?: 40 (or (begin (try-module-linked name)) (try-module-autoload name) ...) ?: 41* [try-module-autoload (gnucash report iframe-url)] ?: 42 (let* (# # # #) (resolve-module dir-hint-module-name #f) (and # #)) ... ?: 43 (letrec ((load-file #)) (dynamic-wind (lambda () #) (lambda () #) ...) ...) ?: 44* [dynamic-wind #<procedure #f ()> #<procedure #f ()> #<procedure #f ()>] ?: 45* [#<procedure #f ()>] ?: 46* (let ((file #)) (cond (# => #) (# => #))) ?: 47 [#<procedure #f #> "/usr/share/gnucash/guile-modules/gnucash/report/iframe$ ?: 48 [load-file #<primitive-procedure primitive-load> ...] ?: 49* [save-module-excursion #<procedure #f ()>] ?: 50 (let (# #) (dynamic-wind # thunk #)) ?: 51 [dynamic-wind #<procedure #f ()> #<procedure #f ()> #<procedure #f ()>] ?: 52* [#<procedure #f ()>] ?: 53* [primitive-load "/usr/share/gnucash/guile-modules/gnucash/report/iframe-ur$ In /usr/share/gnucash/guile-modules/gnucash/report/iframe-url.scm: 7: 54* [require format] In unknown file: ?: 55 (cond ((not feature) (set! *catalog* #f)) ((slib:provided? feature)) ...) ... ?: 56 [scm-error misc-error #f ...] <unnamed port>: In procedure scm-error in expression (scm-error (quote misc-error) #f ...): <unnamed port>: slib:require unsupported feature format
weeve, your pack (?), your bump, your fix ?
This is one of those packages that doesn't really have a maintainer, but since I fixed an issue with it once so i could get something on sparc to work, I normally end up with the update bugs :-P As I'm currently away on business, it may be some time before I can do anything about this. Do the more recent versions of gnumeric still depend on slib, and is slib a hard or variable dependency? If this is a major issue, I can mask it until I have more time to look into it.
probably should mask slib-3.1.1 to prevent problems until it is fixed .. seems to be a fixed dependency at this point for gnucash .. I have not noticed if gnumeric has a problem with it.
In testing I've found that if you emerge gnucash once slib-3.1.1 is on the system, it seems to work (or for sparc at least). Is gnucash crashing before you get a chance to do anything?
it starts to load and then the error occurs and it crashes .. I checked by putting on a machine that never had gnucash installed before .. same problem reverting to 2.4.6 fixed the problem ,,. anything else you would like me to try?
Yeah, when I tested on the sparc, I didn't have a previous version of gnucash installed. I've masked slib-3.1.1 for now on x86 until i get a chance to further investigate when i get back (sometime after 05 March 2004).
Just confirming that slib-3.1.1 is exactly what's breaking gnucash at runtime (downgrading to 2.4.6 fixes). It's ~x86 masked, but not hard masked.
Files missing from slib-3.1.1 are: fmtdoc.txi format.scm formatst.scm these are in slib-2.4.6 .. I'll check if adding these from the 2.4.6 zip to the 3.1.1 zip makes a difference... well we get the following errors now .. (note run from comand line so we see the error feed back) Gdk-WARNING **: shmat failed: error 22 (Invalid argument) Backtrace: In unknown file: ... ?: 37 (begin (if # #) (make-modules-in # full-name)) ?: 38* (if (or # #) (try-load-module name)) ?: 39 [try-load-module (gnucash report iframe-url)] ?: 40 (or (begin (try-module-linked name)) (try-module-autoload name) ...) ?: 41* [try-module-autoload (gnucash report iframe-url)] ?: 42 (let* (# # # #) (resolve-module dir-hint-module-name #f) (and # #)) ... ?: 43 (letrec ((load-file #)) (dynamic-wind (lambda () #) (lambda () #) ...) ...) ?: 44* [dynamic-wind #<procedure #f ()> #<procedure #f ()> #<procedure #f ()> ] ?: 45* [#<procedure #f ()>] ?: 46* (let ((file #)) (cond (# => #) (# => #))) ?: 47 [#<procedure #f #> "/usr/share/gnucash/guile-modules/gnucash/report/if rame$ ?: 48 [load-file #<primitive-procedure primitive-load> ...] ?: 49* [save-module-excursion #<procedure #f ()>] ?: 50 (let (# #) (dynamic-wind # thunk #)) ?: 51 [dynamic-wind #<procedure #f ()> #<procedure #f ()> #<procedure #f ()> ] ?: 52* [#<procedure #f ()>] ?: 53* [primitive-load "/usr/share/gnucash/guile-modules/gnucash/report/ifram e-ur$ In /usr/share/gnucash/guile-modules/gnucash/report/iframe-url.scm: 7: 54* [require format] In unknown file: ?: 55 (cond ((not feature) (set! *catalog* #f)) ((slib:provided? feature)) . ..) ... ?: 56 [scm-error misc-error #f ...] <unnamed port>: In procedure scm-error in expression (scm-error (quote misc-erro r) #f ...): <unnamed port>: slib:require unsupported feature format ... Nope still bombs .. something more subtile is going on ..
I had just added the x86 hard mask (via -x86 keyword) earlier today, so that may be why yours is only at ~x86. On an additional note, I can now reproduce on sparc.
Looks like the bug is open with gnucash but no action yet. See http://bugzilla.gnome.org/show_bug.cgi?id=128701 for their bug.
The following notes may be pertient to the discussion .. as some major changes occurred in version 3.1.1 with respect to "format" http://www-swiss.ai.mit.edu/~jaffer/SLIB.FAQ might need changes at the gnucash or guile level ...
*** Bug 41883 has been marked as a duplicate of this bug. ***
This is resolved in GnuCash 1.8.9; see Bug#47679.
Jason, can you mark this as resolved. 1.8.9 is upstream. Seems that the fault was caused by slib no longer is re-entrant on some parts of the api. Fix is in the 1.8.9. That explains the shmem fault in gdk.
Marking as resolved.
Unfortunately the ``mask'' of x86 by -x86 keyword remains :(
