1.8.1 failure

[Daniel Berger]

Solaris 9
gcc 3.3.1

yaml2byte.c: In function `bytestring_alloc':
yaml2byte.c:42: warning: integer constant is too large for "long" type
yaml2byte.c: In function `bytestring_append':
yaml2byte.c:56: warning: integer constant is too large for "long" type
yaml2byte.c: In function `bytestring_extend':
yaml2byte.c:89: warning: integer constant is too large for "long" type
yaml2byte.c:90: warning: integer constant is too large for "long" type
yaml2byte.c:92: error: `YAMLBYTE_ANCHOR' undeclared (first use in this
function)
yaml2byte.c:92: error: (Each undeclared identifier is reported only once
yaml2byte.c:92: error: for each function it appears in.)
yaml2byte.c:96: error: `YAMLBYTE_ALIAS' undeclared (first use in this
function)
yaml2byte.c: In function `syck_yaml2byte_handler':
yaml2byte.c:134: error: `YAMLBYTE_ANCHOR' undeclared (first use in this
function)
yaml2byte.c:139: error: `YAMLBYTE_TRANSFER' undeclared (first use in
this function)
yaml2byte.c:153: error: `YAMLBYTE_SCALAR' undeclared (first use in this
function)
yaml2byte.c:163: error: `YAMLBYTE_CONTINUE' undeclared (first use in
this function)
yaml2byte.c:172: error: `YAMLBYTE_NEWLINE' undeclared (first use in this
function)
yaml2byte.c:176: error: `YAMLBYTE_NULLCHAR' undeclared (first use in
this function)
yaml2byte.c:187: error: `YAMLBYTE_SEQUENCE' undeclared (first use in
this function)
yaml2byte.c:194: error: `YAMLBYTE_END_BRANCH' undeclared (first use in
this function)
yaml2byte.c:197: error: `YAMLBYTE_MAPPING' undeclared (first use in this
function)
make[1]: *** [yaml2byte.o] Error 1
make[1]: Leaving directory `/usr/local/src/ruby-1.8.1/ext/syck'
make: *** [all] Error 1

I see an include for yamlbyte.h in yaml2byte.c. Is this a dynamically
generated file of some sort? Because I don't see it.

/usr/local/src/ruby-1.8.1/config.h:16:1: warning: "_FILE_OFFSET_BITS"
redefined

Wasn't a patch submitted to fix this warning already? I seem to
remember it being submitted on the core list.

Regards,

Dan

 

[Yukihiro Matsumoto]

Hi,

In message "1.8.1 failure"

|yaml2byte.c: In function `bytestring_alloc':
|yaml2byte.c:42: warning: integer constant is too large for "long" type
|yaml2byte.c: In function `bytestring_append':

Thank you for the report. I just put preview2 at

ftp://ftp.ruby-lang.org/pub/ruby/1.8/ruby-1.8.1-preview2.tar.gz

md5sum is 067deb137b70d6ca046f6d7e9a84e198

matz.

 

[why the lucky stiff]

yaml2byte.c: In function `bytestring_alloc':
yaml2byte.c:42: warning: integer constant is too large for "long" type

Sorry about that (Matz and others). I was warned of these errors, but the
preview sort of snuck up all the sudden. Indeed, these should be fixed
presently.

Nice work on this preview release, matz. And everyone else out there: please
report Syck or YAML-related errors straightway to me.

_why

 

[Yukihiro Matsumoto]

Hi,

In message "Re: 1.8.1 failure"

|Sorry about that (Matz and others). I was warned of these errors, but the
|preview sort of snuck up all the sudden. Indeed, these should be fixed
|presently.

It's the purpose of previews, but I still should have told you the
preview release schedule. Sorry. We will have a few previews soon.
Then the final 1.8.1 release will be in two weeks or so.

matz.

 

[Corinna Vinschen]

Hi,

In message "1.8.1 failure"

|yaml2byte.c: In function `bytestring_alloc':
|yaml2byte.c:42: warning: integer constant is too large for "long" type
|yaml2byte.c: In function `bytestring_append':

Thank you for the report. I just put preview2 at

ftp://ftp.ruby-lang.org/pub/ruby/1.8/ruby-1.8.1-preview2.tar.gz

Testing on current Cygwin, gcc 3.3.1, most of the ext subdirectory
doesn't get build. I configured with

--enable-shared --with-X11-dir=/usr/X11R6

The configure log doesn't show anything abnormal. The following `make'
runs fine until it enters the ext subdirectory. There it builds only
the first three directories, bigdecimal, curses and dbm. For each
following directory, the output is:

compiling digest
make[1]: Entering directory `/home/corinna/src/ruby-1.8.1/ext/digest'
make[1]: Nothing to be done for `all'.
make[1]: Leaving directory `/home/corinna/src/ruby-1.8.1/ext/digest'
compiling digest/md5
make[1]: Entering directory `/home/corinna/src/ruby-1.8.1/ext/digest/md5'
make[1]: Nothing to be done for `all'.
make[1]: Leaving directory `/home/corinna/src/ruby-1.8.1/ext/digest/md5'
[...]

without building anything in all these directories. Up to that point,
no error message or warning has been emited. After leaving ext, the
remaining build of ruby.exe and rubyw.exe in the top level directory
succeeds again.


Corinna

 

[gabriele renzi]

il Fri, 31 Oct 2003 08:01:58 +0900, (e-mail address removed) (Yukihiro
Matsumoto) ha scritto::

Thank you for the report. I just put preview2 at

ftp://ftp.ruby-lang.org/pub/ruby/1.8/ruby-1.8.1-preview2.tar.gz

md5sum is 067deb137b70d6ca046f6d7e9a84e198

matz.

A little note, problem running test/runner.rb:

for some strange reason, I dind'nt had a mapping from
localhost_hostname to 127.0.0.1.
running test/runner.rb hangs up with this error:

home/nickel/lib/ruby/1.8/drb/drb.rb:693:in `open':
druby://UltimaThule:36121 -
#<SocketError: getaddrinfo: Temporary failure in name resolution>
(DRb:RbConnError)
from /home/nickel/lib/ruby/1.8/drb/drb.rb:686:in `each'
from /home/nickel/lib/ruby/1.8/drb/drb.rb:686:in `open'
from /home/nickel/lib/ruby/1.8/drb/drb.rb:1072:in `initialize'
from /home/nickel/lib/ruby/1.8/drb/drb.rb:1055:in `new'
from /home/nickel/lib/ruby/1.8/drb/drb.rb:1055:in `open'
from /home/nickel/lib/ruby/1.8/drb/drb.rb:1002:in
`method_missing'
from /home/nickel/lib/ruby/1.8/drb/extserv.rb:16:in
`initialize'
from /usr/src/ruby-1.8.1/test/drb/ut_array.rb:12:in `new'
from /usr/src/ruby-1.8.1/test/drb/ut_array.rb:12


I understand the error, but I don't think this should freeze the
script..

Anyway, after a little change to etc/hosts everything works fine[1].

Another thing: why 'make test' does'nt run the test scripts ?




[1]
actually, tests run fine even with
CFLAGS= -march=athlon-xp -O3 -pipe -fomit-frame-pointer and
gcc version 3.2 (Mandrake Linux 9.0 3.2-1mdk)
obviously, not great enhancement over -O2