2.6.0-test10: Badness in local_bh_enable at kernel/softirq.c:121

2.6.0-test10: Badness in local_bh_enable at kernel/softirq.c:121

[Christian Schlittchen]

When trying to establish a ppp/pppoe connection I get the following
and the connection fails:

Badness in local_bh_enable at kernel/softirq.c:121
Call Trace:
[<c011feac>] local_bh_enable+0x8c/0x90
[<e096ccae>] ppp_sync_push+0x6e/0x1a0 [ppp_synctty]
[<c015cdc0>] __lookup_hash+0x70/0xd0
[<e096c651>] ppp_sync_wakeup+0x31/0x70 [ppp_synctty]
[<c0207b79>] pty_unthrottle+0x59/0x60
[<c02043ba>] check_unthrottle+0x3a/0x40
[<c0204463>] n_tty_flush_buffer+0x13/0x60
[<c0207f6d>] pty_flush_buffer+0x6d/0x70
[<c0200c0e>] do_tty_hangup+0x3fe/0x460
[<c0202246>] release_dev+0x656/0x6b0
[<c014010b>] zap_pmd_range+0x4b/0x70
[<c0140173>] unmap_page_range+0x43/0x70
[<c01656a2>] dput+0x22/0x220
[<c0202600>] tty_release+0x0/0x70
[<c020262a>] tty_release+0x2a/0x70
[<c014f468>] __fput+0x118/0x130
[<c014d9d9>] filp_close+0x59/0x90
[<c011d70c>] put_files_struct+0x5c/0xd0
[<c011e33f>] do_exit+0x15f/0x3b0
[<c011e66b>] do_group_exit+0x7b/0xc0
[<c010935b>] syscall_call+0x7/0xb

This is 100% reproducible with any kernel version since at least 2.5.71
I tried. I tried to recompile the kernel with different settings and
compilers, but the error stays the same.

I can provide more information if needed.

Regards, Christian

Re: 2.6.0-test10: Badness in local_bh_enable at kernel/softirq.c:121

[Andrew Morton]

Christian Schlittchen <corwin@amber.kn-bremen.de> wrote:
>
> 
> When trying to establish a ppp/pppoe connection I get the following
> and the connection fails:
> 
> Badness in local_bh_enable at kernel/softirq.c:121
> Call Trace:
> [<c011feac>] local_bh_enable+0x8c/0x90
> [<e096ccae>] ppp_sync_push+0x6e/0x1a0 [ppp_synctty]
> [<c015cdc0>] __lookup_hash+0x70/0xd0
> [<e096c651>] ppp_sync_wakeup+0x31/0x70 [ppp_synctty]
> [<c0207b79>] pty_unthrottle+0x59/0x60
> [<c02043ba>] check_unthrottle+0x3a/0x40
> [<c0204463>] n_tty_flush_buffer+0x13/0x60
> [<c0207f6d>] pty_flush_buffer+0x6d/0x70
> [<c0200c0e>] do_tty_hangup+0x3fe/0x460

The warning is a pest, and is due to do_tty_hangup() bogusly disabling
interrupts in the hope that it does something useful.  It needs to be fixed
up.

But it is unrelated to the PPP failure.  I'm afraid it is so long since I
used PPP and pppd that I cannot suggest how you should set about gathering
extra info on that.

Re: 2.6.0-test10: Badness in local_bh_enable at kernel/softirq.c:121

[Pete Clements]

Quoting Andrew Morton
  > Christian Schlittchen <corwin@amber.kn-bremen.de> wrote:
  > >
  > > 
  > > When trying to establish a ppp/pppoe connection I get the following
  > > and the connection fails:
  > > 
  > > Badness in local_bh_enable at kernel/softirq.c:121
  > > Call Trace:
  > > [<c011feac>] local_bh_enable+0x8c/0x90
  > > [<e096ccae>] ppp_sync_push+0x6e/0x1a0 [ppp_synctty]
  > > [<c015cdc0>] __lookup_hash+0x70/0xd0
  > > [<e096c651>] ppp_sync_wakeup+0x31/0x70 [ppp_synctty]
  > > [<c0207b79>] pty_unthrottle+0x59/0x60
  > > [<c02043ba>] check_unthrottle+0x3a/0x40
  > > [<c0204463>] n_tty_flush_buffer+0x13/0x60
  > > [<c0207f6d>] pty_flush_buffer+0x6d/0x70
  > > [<c0200c0e>] do_tty_hangup+0x3fe/0x460
  > 
  > The warning is a pest, and is due to do_tty_hangup() bogusly disabling
  > interrupts in the hope that it does something useful.  It needs to be fixed
  > up.
  > 
  > But it is unrelated to the PPP failure.  I'm afraid it is so long since I
  > used PPP and pppd that I cannot suggest how you should set about gathering
  > extra info on that.
  > 
What network card? Multi-cards? Is the card detected?

Around 2.5.71 the 3c509 init/detect changed such that the
second card is not detected.  Will get the badness when attempting
to bring up ppp/ppoe on a non-existent card.  

The 3c509 multi card breakage still exists at test10. I replace the
driver with 2.5.7x version to get my pppoe.
-- 
Pete Clements