On Fri, Aug 01, 2014 at 08:54:07AM -0700, Greg KH wrote: > On Tue, Jun 24, 2014 at 03:52:05PM +0200, Antoine Schweitzer-Chaput wrote: > > The code which was only reachable using these options is also removed. > > > > Signed-off-by: Antoine Schweitzer-Chaput <antoine@xxxxxxxxxxxxxxxxxxxx> > > --- > > drivers/staging/rtl8192u/Makefile | 3 - > > drivers/staging/rtl8192u/r8192U_core.c | 482 ++------------------------------- > > 2 files changed, 16 insertions(+), 469 deletions(-) > > > > diff --git a/drivers/staging/rtl8192u/Makefile b/drivers/staging/rtl8192u/Makefile > > index eefc657..703c150 100644 > > --- a/drivers/staging/rtl8192u/Makefile > > +++ b/drivers/staging/rtl8192u/Makefile > > @@ -6,9 +6,6 @@ ccflags-y += -O2 > > ccflags-y += -DCONFIG_FORCE_HARD_FLOAT=y > > ccflags-y += -DJACKSON_NEW_8187 -DJACKSON_NEW_RX > > ccflags-y += -DTHOMAS_BEACON -DTHOMAS_TASKLET -DTHOMAS_SKB -DTHOMAS_TURBO > > -#ccflags-y += -DUSB_TX_DRIVER_AGGREGATION_ENABLE > > -#ccflags-y += -DUSB_RX_AGGREGATION_SUPPORT > > -ccflags-y += -DUSE_ONE_PIPE > > Why delete this last one if it is being set? Since it is always set, and not modifiable anywhere, it allows removing the code in r8192U_core.c under #ifndef USE_ONE_PIPE > Did you test this change on hardware? No, I didn't have the chance. I however compared the generated code after preprocessing (using gcc -E -P) with and without this change, and the only differences are the 3 removed functions (dump_eprom, txqueue2outpipe, and GetRxPacketShiftBytes819xUsb). Antoine _______________________________________________ devel mailing list devel@xxxxxxxxxxxxxxxxxxxxxx http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
