Ath5k compile error in wireless-2.6.git

John,

I just checked out the wireless-2.6.git tree from Kernel.org to test the new ath5k driver. I get the following compile error on a 'vanilla' git-clone.

CC [M] drivers/net/wireless/ath5k/base.o
In file included from drivers/net/wireless/ath5k/base.h:49, from drivers/net/wireless/ath5k/base.c:58:
drivers/net/wireless/ath5k/ath5k.h:213: error: redeclaration of enumerator 'MODE_ATHEROS_TURBO'
include/net/mac80211.h:105: error: previous definition of 'MODE_ATHEROS_TURBO' was here drivers/net/wireless/ath5k/ath5k.h:215: error: redeclaration of enumerator 'MODE_ATHEROS_TURBOG'
include/net/mac80211.h:107: error: previous definition of 'MODE_ATHEROS_TURBOG' was here drivers/net/wireless/ath5k/base.c: In function 'ath_configure_filter':
drivers/net/wireless/ath5k/base.c:1419: error: 'FIF_PROMISC_IN_BSS' undeclared (first use in this function) drivers/net/wireless/ath5k/base.c:1419: error: (Each undeclared identifier is reported only once drivers/net/wireless/ath5k/base.c:1419: error: for each function it appears in.)
drivers/net/wireless/ath5k/base.c:1419: error: 'FIF_ALLMULTI' undeclared (first use in this function) drivers/net/wireless/ath5k/base.c:1419: error: 'FIF_FCSFAIL' undeclared (first use in this function) drivers/net/wireless/ath5k/base.c:1419: error: 'FIF_PLCPFAIL' undeclared (first use in this function) drivers/net/wireless/ath5k/base.c:1419: error: 'FIF_CONTROL' undeclared (first use in this function) drivers/net/wireless/ath5k/base.c:1419: error: 'FIF_OTHER_BSS' undeclared (first use in this function) drivers/net/wireless/ath5k/base.c:1419: error: 'FIF_BCN_PRBRESP_PROMISC' undeclared (first use in this function) drivers/net/wireless/ath5k/base.c: At top level:
drivers/net/wireless/ath5k/base.c:1479: warning: 'enum set_key_cmd' declared inside parameter list drivers/net/wireless/ath5k/base.c:1479: warning: its scope is only this definition or declaration, which is probably not what you want drivers/net/wireless/ath5k/base.c:1477: error: parameter 2 ('cmd') has incomplete type
drivers/net/wireless/ath5k/base.c:1583: error: unknown field 'start' specified in initializer drivers/net/wireless/ath5k/base.c:1584: warning: initialization from incompatible pointer type drivers/net/wireless/ath5k/base.c:1589: error: unknown field 'configure_filter' specified in initializer drivers/net/wireless/ath5k/base.c:1589: warning: initialization from incompatible pointer type drivers/net/wireless/ath5k/base.c:1590: warning: initialization from incompatible pointer type
make[4]: *** [drivers/net/wireless/ath5k/base.o] Error 1 make[3]: *** [drivers/net/wireless/ath5k] Error 2
make[2]: *** [drivers/net/wireless] Error 2
make[1]: *** [drivers/net] Error 2

Maybe you can shed some light?

John,

I just checked out the wireless-2.6.git tree from Kernel.org to test the new ath5k driver. I get the following compile error on a 'vanilla' git-clone.

Quoted 5 lines · John Daiker

John,

I just checked out the wireless-2.6.git tree from Kernel.org to test the new ath5k driver. I get the following compile error on a 'vanilla' git-clone.

Please use the 'everything' branch for development.

git checkout everything

Quoted 10 lines · John W. Linville

Quoted 5 lines · John Daiker

John,

I just checked out the wireless-2.6.git tree from Kernel.org to test the new ath5k driver. I get the following compile error on a 'vanilla' git-clone.

Please use the 'everything' branch for development.

git checkout everything

Hmmm...maybe this instead:

git checkout -b everything origin/everything

Quoted 11 lines · John W. Linville

Quoted 10 lines · John W. Linville

Quoted 5 lines · John Daiker

John,

I just checked out the wireless-2.6.git tree from Kernel.org to test the new ath5k driver. I get the following compile error on a 'vanilla' git-clone.

Please use the 'everything' branch for development.

git checkout everything

Approach 1

Quoted 2 lines

Hmmm...maybe this instead:

Quoted 1 line

git checkout -b everything origin/everything

Approach 2

What's fun is that the above is approach 1 is actually necessary on PowerPC. It seems git for powerpc still uses the old branch stuff. For example I can't use approach 2. I'll add this to the wiki for ath5k, just a head up in case there's some PowerPC users out there.

Quoted 21 lines · Luis R. Rodriguez

Quoted 11 lines · John W. Linville

Quoted 10 lines · John W. Linville

Quoted 5 lines · John Daiker

John,

I just checked out the wireless-2.6.git tree from Kernel.org to test the new ath5k driver. I get the following compile error on a 'vanilla' git-clone.

Please use the 'everything' branch for development.

git checkout everything

Approach 1

Quoted 3 lines

Hmmm...maybe this instead:

git checkout -b everything origin/everything

Approach 2

Quoted 4 lines

What's fun is that the above is approach 1 is actually necessary on PowerPC. It seems git for powerpc still uses the old branch stuff. For example I can't use approach 2. I'll add this to the wiki for ath5k, just a head up in case there's some PowerPC users out there.

I think this has more to do with the git version (i.e. release) than the architecture. "Approach 1" worked everywhere for a long time, while I only became aware of "Approach 2" a few months ago.

Quoted 3 lines · John W. Linville

I think this has more to do with the git version (i.e. release) than the architecture. "Approach 1" worked everywhere for a long time, while I only became aware of "Approach 2" a few months ago.

On my x86 box I have git 1.5.2.5
On my PPC box have git 1.5.3.4

And no dice on the PowerPC box. Maybe I have to do a new pull with the new release.. I'll try.