Don't assume that both iowrite32() and iowrite32be() return values in
the _OS_REG_WRITE definition. _OS_REG_WRITE discards the value anyway.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@4061 0192ed92-7a03-0410-a25b-9323aeb14dbd
32-bit PowerPC only gets iowrite32be() and ioread32be() in version
2.6.20. This applies only to the merged "powerpc" architecture
identified by CONFIG_PPC_MERGE. The old "ppc" architecture goes away in
Linux 2.6.27, so CONFIG_PPC_MERGE is no longer valid.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@4029 0192ed92-7a03-0410-a25b-9323aeb14dbd
It's obsolete and useless now, when the HAL sources are open. It also
misuses ath_hal_debug in a way incompatible with the use of
ath_hal_debug inside the HAL.
The mmiotrace facility is available in the kernel and doesn't require
any changes to the sources.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@4020 0192ed92-7a03-0410-a25b-9323aeb14dbd
Provide forward declaration of struct ath_hal in ah_osdep.h. Implement
OS_GETUPTIME (the implementation is quite primitive, as it's only used
once for some obscure condition in 5212 chips). Implement HALDEBUG()
and ath_hal_ether_sprintf().
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@4019 0192ed92-7a03-0410-a25b-9323aeb14dbd
Use correct shifts for the rate data. Remove excessive debug statements
that would have to be extended even further to deal with AR5416.
The code has been taken from the FreeBSD driver. The original coding
style has been preserved.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@4015 0192ed92-7a03-0410-a25b-9323aeb14dbd
It's better to keep initialization in one place to simplify conversion
to the new style netdev ops.
Declare ieee80211_ioctl() in ieee80211_linux.h next to other
ieee80211_ioctl* functions defined in ieee80211_wireless.c.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@4000 0192ed92-7a03-0410-a25b-9323aeb14dbd
Use spin_lock_irqsave_nested() instead of spin_lock_irqsave(). Use a
separate counter for the locking subclass, so that we don't exceed the
maximal value of 8. We have 10 queues, but inactive queues don't get a
locking subclass number.
Provide a replacement for spin_lock_irqsave_nested() for kernels older
than 2.6.19.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@3995 0192ed92-7a03-0410-a25b-9323aeb14dbd
Check that AH_PRIVATE(_ah)->ah_eepromDetach is not NULL before calling
it. This fixes kernel oops in case if the attach function for the chip
fails before EEPROM support initialization and calls the corresponding
detach function.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@3991 0192ed92-7a03-0410-a25b-9323aeb14dbd
When the compiler tries inlining static functions, it can notice that
some variables may not be initialized in the functions being inlined.
Provide initialization for output variables in all branches of
execution, no matter how anomalous. Add asserts in some cases.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@3988 0192ed92-7a03-0410-a25b-9323aeb14dbd
gcc warns that "array subscript is above array bounds" with -O3 since
pvar is assigned before _i is checked.
Eliminate _i, use a new ppvar variable, which is a pointer to the
reference table. Terminate reference tables with NULL instead of using
ARRAY_SIZE. Fix types of reference tables to match those actually used
by the code.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@3987 0192ed92-7a03-0410-a25b-9323aeb14dbd
Linux 2.6.30 doesn't define IRQ_NONE as a macro. Assume irqreturn_t,
IRQ_NONE and IRQ_HANDLED to be present on Linux 2.6.29 and newer.
git-svn-id: http://madwifi-project.org/svn/madwifi/trunk@3985 0192ed92-7a03-0410-a25b-9323aeb14dbd