2.4.8 breaks ATM

2.4.8 breaks ATM

[Renaud Guérin]

The following part of the 2.4.8 patch seems to be a typo (">>" looks more 
logical than ">" if vci_bit is a bitfield), and broke ATM for me:

--- v2.4.7/linux/net/atm/common.c       Tue Jul  3 17:08:22 2001
+++ linux/net/atm/common.c      Sun Aug  5 13:12:41 2001
@@ -210,7 +210,7 @@
 
        if ((vpi != ATM_VPI_UNSPEC && vpi != ATM_VPI_ANY &&
            vpi >> dev->ci_range.vpi_bits) || (vci != ATM_VCI_UNSPEC &&
-           vci != ATM_VCI_ANY && vci >> dev->ci_range.vci_bits))
+           vci != ATM_VCI_ANY && vci > dev->ci_range.vci_bits))
                return -EINVAL;
        if (vci > 0 && vci < ATM_NOT_RSV_VCI && 
!capable(CAP_NET_BIND_SERVICE))
                return -EPERM;


I simply reverted the patch and it works again.


-- 
-----------------------------------------------
Renaud Guerin
rguerin@free.fr - guerinre@utt.fr
Génie des Systèmes d'Information et de Décision
Université de Technologie de Troyes (France)
-----------------------------------------------

Re: 2.4.8 breaks ATM

[Mitchell Blank Jr]

Renaud Guérin wrote:
> The following part of the 2.4.8 patch seems to be a typo (">>" looks more 
> logical than ">" if vci_bit is a bitfield), and broke ATM for me:

Yes, the patch is 100% wrong and should be reverted.  It had been sent to
the ATM maintainers and was rejected, but someone else must have submitted
it upstream since it magically appeared in the mainstream kernel.  It's
already been reverted in both 2.4.9pre1 and 2.4.8ac1.

-Mitch