configure fails at getaddrinfo() check on KAME

[suzsuz]

BPO	934635
Nosy	@hyeshik

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = 'https://github.com/hyeshik'
closed_at = <Date 2004-04-14.08:15:47.000>
created_at = <Date 2004-04-14.00:26:13.000>
labels = ['build']
title = 'configure fails at getaddrinfo() check on KAME'
updated_at = <Date 2004-04-14.08:15:47.000>
user = 'https://bugs.python.org/suzsuz'

bugs.python.org fields:

activity = <Date 2004-04-14.08:15:47.000>
actor = 'hyeshik.chang'
assignee = 'hyeshik.chang'
closed = True
closed_date = None
closer = None
components = ['Build']
creation = <Date 2004-04-14.00:26:13.000>
creator = 'suzsuz'
dependencies = []
files = []
hgrepos = []
issue_num = 934635
keywords = []
message_count = 3.0
messages = ['20500', '20501', '20502']
nosy_count = 2.0
nosy_names = ['hyeshik.chang', 'suzsuz']
pr_nums = []
priority = 'normal'
resolution = 'fixed'
stage = None
status = 'closed'
superseder = None
type = None
url = 'https://bugs.python.org/issue934635'
versions = ['Python 2.4']

[suzsuz]

[Phenomena]
Python's configure script fails at getaddrinfo() check
on KAME-SNAP-freebsd4-20040412, although it supports
getaddrinfo() properly.

configure:14128: checking getaddrinfo bug
configure:14230: cc -o conftest -O -pipe
-D_THREAD_SAFE -DTHREAD_STACK_SIZE=0x20000 -pthread
conftest.c -L/usr/local/v6/lib -linet6 -lutil >&5
configure:14233: $? = 0
configure:14235: ./conftest
configure:14238: $? = 1
configure: program exited with status 1
configure: failed program was:
...

[reason]
The testing program to verify getaddrinfo() in
configure script assumes that there is only one
SOCK_STREAM protocol (normally TCP).

Since KAME supports SCTP as well as TCP, this
assumption is not true; getaddrinfo() returns two
protocols for the same socket type, which lead to a
unexpected result in this testing program.

[Solution]
Just specify the protocol in the testing program (like
the following patch) so that new protocol support
should not lead to such configure failure.

     hints.ai_family = AF_UNSPEC;
     hints.ai_flags = passive ? AI_PASSIVE : 0;
     hints.ai_socktype = SOCK_STREAM;
+    hints.ai_protocol = IPPROTO_TCP;
     if ((gaierr = getaddrinfo(NULL, "54321", &hints,
&aitop)) != 0) {
       (void)gai_strerror(gaierr);
       goto bad;

[hyeshik]

Logged In: YES
user_id=55188

Looks correct. I'll apply this in CVS, soon.

[hyeshik]

Logged In: YES
user_id=55188

Checked in

(head)
configure 1.443
configure.in 1.454
Misc/NEWS 1.966

(2.3 maint)
configure 1.416.4.13
configure.in 1.427.4.12
Misc/NEWS 1.831.4.99

Thank you for the kind description. :)