Messages in this thread | | | Date | Thu, 17 Apr 2008 15:39:29 -0400 | From | Jeff Garzik <> | Subject | Re: [2.6 patch] xtensa: don't offer ATA |
| |
Adrian Bunk wrote: > Trying to build libata for xtensa fails with the following error: > > <-- snip --> > > ... > CC [M] drivers/ata/libata-core.o > /home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ata/libata-core.c: In function 'ata_devchk': > /home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ata/libata-core.c:1034: error: implicit declaration of function 'iowrite8' > /home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ata/libata-core.c:1043: error: implicit declaration of function 'ioread8' > /home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ata/libata-core.c: In function 'ata_data_xfer': > /home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ata/libata-core.c:5027: error: implicit declaration of function 'ioread16_rep' > /home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ata/libata-core.c:5029: error: implicit declaration of function 'iowrite16_rep' > /home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ata/libata-core.c:5037: error: implicit declaration of function 'ioread16' > /home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ata/libata-core.c:5041: error: implicit declaration of function 'iowrite16' > /home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ata/libata-core.c: In function 'ata_wait_register': > /home/bunk/linux/kernel-2.6/git/linux-2.6/drivers/ata/libata-core.c:7700: error: implicit declaration of function 'ioread32' > make[3]: *** [drivers/ata/libata-core.o] Error 1 > > <-- snip --> > > Signed-off-by: Adrian Bunk <bunk@kernel.org> > > --- > aa579e7ab065e41a9d1e27e1f4b7af6b0508018c diff --git a/drivers/ata/Kconfig b/drivers/ata/Kconfig > index 25aba69..f34e9d1 100644 > --- a/drivers/ata/Kconfig > +++ b/drivers/ata/Kconfig > @@ -6,7 +6,7 @@ menuconfig ATA > tristate "Serial ATA (prod) and Parallel ATA (experimental) drivers" > depends on HAS_IOMEM > depends on BLOCK > - depends on !(M32R || M68K) || BROKEN > + depends on !(M32R || M68K || XTENSA) || BROKEN > depends on !SUN4 || BROKEN > select SCSI
XTENSA offers HAS_IOMEM but not iomem functions? That seems like the problem to me.
Overall, we should not be adding specific arches to an exclusion list, we should defining HAS_IOFOO or MISSING_IOFOO or another symbol in the appropriate arch.
Jeff
| |