Mailinglist Archive: opensuse-bugs (6595 mails)

< Previous Next >
[Bug 532575] latest Factory update to aria2 broke zypper/YaST
  • From: bugzilla_noreply@xxxxxxxxxx
  • Date: Thu, 20 Aug 2009 22:40:06 -0600
  • Message-id: <20090821044006.13F63CC7CD@xxxxxxxxxxxxxxxxxxxxxx>
http://bugzilla.novell.com/show_bug.cgi?id=532575

User tatsuhiro.t@xxxxxxxxx added comment
http://bugzilla.novell.com/show_bug.cgi?id=532575#c9





--- Comment #9 from Tatsuhiro Tsujikawa <tatsuhiro.t@xxxxxxxxx> 2009-08-20
22:40:00 MDT ---
I fixed this bug in r1498 and yes this is a regression since 1.5.0.

When connecting to the first resolved address is failed, then aria2 don't try
another one. This is the cause of bug.

In strace log, aria2 seems to connect to all resolved address, but it is not.

connect(5, {sa_family=AF_INET6, sin6_port=htons(0), inet_pton(AF_INET6,
"2a02:6b8:0:201::1", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, 28) = -1
ENETUNREACH (Network is unreachable)


This failed - okay. Now, aria2c connects to the other addresses, which works,
but it keeps trying and gives up after trying all four addresses:



connect(5, {sa_family=AF_UNSPEC, sa_data="\0\0\0\0\0\0\0\0\0\0\0\0\0\0"}, 16)
=
0
connect(5, {sa_family=AF_INET, sin_port=htons(0),
sin_addr=inet_addr("93.158.155.193")}, 16) = 0
getsockname(5, {sa_family=AF_INET6, sin6_port=htons(58831),
inet_pton(AF_INET6,
"::ffff:83.133.126.38", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, [28])
=
0
connect(5, {sa_family=AF_UNSPEC, sa_data="\0\0\0\0\0\0\0\0\0\0\0\0\0\0"}, 16)
=
0
connect(5, {sa_family=AF_INET, sin_port=htons(0),
sin_addr=inet_addr("77.88.19.68")}, 16) = 0
getsockname(5, {sa_family=AF_INET6, sin6_port=htons(43340),
inet_pton(AF_INET6,
"::ffff:83.133.126.38", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, [28])
=
0
connect(5, {sa_family=AF_UNSPEC, sa_data="\0\0\0\0\0\0\0\0\0\0\0\0\0\0"}, 16)
=
0
connect(5, {sa_family=AF_INET, sin_port=htons(0),
sin_addr=inet_addr("77.88.19.73")}, 16) = 0
getsockname(5, {sa_family=AF_INET6, sin6_port=htons(35944),
inet_pton(AF_INET6,
"::ffff:83.133.126.38", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, [28])
=
0
connect(5, {sa_family=AF_UNSPEC, sa_data="\0\0\0\0\0\0\0\0\0\0\0\0\0\0"}, 16)
=
0
connect(5, {sa_family=AF_INET, sin_port=htons(0),
sin_addr=inet_addr("77.88.19.74")}, 16) = 0
getsockname(5, {sa_family=AF_INET6, sin6_port=htons(43049),
inet_pton(AF_INET6,
"::ffff:83.133.126.38", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, [28])
=
0
close(5) = 0


These attempt to connecting all resolved addresses were done by getaddrinfo()
system call. Not by aria2's application code.
aria2's first failed attempt is way below:

connect(5, {sa_family=AF_INET6, sin6_port=htons(80), inet_pton(AF_INET6,
"2a02:6b8:0:201::1", &sin6_addr), sin6_flowinfo=0, sin6_scope_id=0}, 28) = -1
ENETUNREACH (Network is unreachable)

It is interesting to know why getaddrinfo is connecting to these addresses, but
it is not to related this issue. Anyway, the fix to this regression is in svn
repository and the new release 1.5.2 is about to release.

--
Configure bugmail: http://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.

< Previous Next >
References