Building guile-2.0.11 on a new cygwin 64

Building guile-2.0.11 on a new cygwin 64

[Mike Gran]

Hi-

I started looking at what it would take to build a
Guile 2.0.11 on Cygwin.  The current prepackaged
Guile on Cygwin is 1.8.8.

I used 
gc-7.4.2
gmp-6.1.0
guile-2.0.11
libatomic_ops-7.4.2
libtool-2.4.6
libunistring-0.9.6

I used gc-7.4.2, but, I had to pull in this patch from master.
https://github.com/ivmai/bdwgc/commit/4e24d219d29abb2b111bb9a9ccb13c3f149a855a

I had to patch 
guile-2.0.11/lib/sys_select.in.h
as described in this e-mail
http://lists.gnu.org/archive/html/bug-gnulib/2016-03/msg00057.html

And I had to build guile with the "--without-threads" configure
option.  When I use threads, it hangs in scm_i_pthread_cond_wait.

And with that, it builds and runs and passes the majority of the
"make check" tests, except for these.

FAIL: 00-socket.test: AF_UNIX/SOCK_STREAM: accept 
FAIL: 00-socket.test: AF_UNIX/SOCK_STREAM: accept (bis) 

FAIL: 00-socket.test: AF_INET6/SOCK_STREAM: accept 
FAIL: i18n.test: character mapping: char-locale-upcase Turkish 
FAIL: i18n.test: character mapping: char-locale-downcase Turkish 
FAIL: i18n.test: string mapping: string-locale-upcase Turkish 
FAIL: i18n.test: string mapping: string-locale-downcase Turkish 
FAIL: i18n.test: number->locale-string: French: integer 
FAIL: i18n.test: number->locale-string: French: fraction 
FAIL: i18n.test: number->locale-string: French: fraction, 1 digit 
FAIL: i18n.test: monetary-amount->locale-string: French: integer 
FAIL: i18n.test: monetary-amount->locale-string: French: fraction 


And the standalone test "test-ffi" fails with

ERROR: In procedure dynamic-func: 
ERROR: In procedure dynamic-pointer: Symbol not found: strerror 


I know this information is all too casual to be a either a bug
report or build instructions, but, I wanted to document what I had
before I moved on to other things.  Hopefully it'll be of use
to someone.

Maybe I'll get back to this again soon.

Thanks,
Mike Gran

Re: Building guile-2.0.11 on a new cygwin 64

[Eli Zaretskii]

> Date: Sun, 29 May 2016 16:17:29 +0000 (UTC)
> From: Mike Gran <spk121@yahoo.com>
> 
> And I had to build guile with the "--without-threads" configure
> option.  When I use threads, it hangs in scm_i_pthread_cond_wait.
> 
> And with that, it builds and runs and passes the majority of the
> "make check" tests, except for these.
> 
> FAIL: 00-socket.test: AF_UNIX/SOCK_STREAM: accept 
> FAIL: 00-socket.test: AF_UNIX/SOCK_STREAM: accept (bis) 
> 
> FAIL: 00-socket.test: AF_INET6/SOCK_STREAM: accept 
> FAIL: i18n.test: character mapping: char-locale-upcase Turkish 
> FAIL: i18n.test: character mapping: char-locale-downcase Turkish 
> FAIL: i18n.test: string mapping: string-locale-upcase Turkish 
> FAIL: i18n.test: string mapping: string-locale-downcase Turkish 
> FAIL: i18n.test: number->locale-string: French: integer 
> FAIL: i18n.test: number->locale-string: French: fraction 
> FAIL: i18n.test: number->locale-string: French: fraction, 1 digit 
> FAIL: i18n.test: monetary-amount->locale-string: French: integer 
> FAIL: i18n.test: monetary-amount->locale-string: French: fraction 
> 
> 
> And the standalone test "test-ffi" fails with
> 
> ERROR: In procedure dynamic-func: 
> ERROR: In procedure dynamic-pointer: Symbol not found: strerror 
> 
> 
> I know this information is all too casual to be a either a bug
> report or build instructions, but, I wanted to document what I had
> before I moved on to other things.  Hopefully it'll be of use
> to someone.
> 
> Maybe I'll get back to this again soon.

I'm surprised you have almost the same problems in a Cygwin build
that I had in the MinGW build (see several discussions back in June
2013 and some more in 2014).  Cygwin is much more Posix compatible, so
at least the i18n tests should have passed.  But maybe I'm missing
something; if you look at the problems I reported back then, perhaps
you will be able to uncover the root causes for these problems.