6e996ee8e7
Update the AT91 MMC driver to use the generic GPIO calls instead of the AT91-specific calls; and to request (and release) those GPIO signals. That required updating the probe() fault cleanup codepaths. Now there is a single sequence for freeing resources, in reverse order of their allocation. Also that code uses use dev_*() for messaging, and has less abuse of KERN_ERR. Likewise with updating remove() cleanup. This had to free the GPIOs, and while adding that code I noticed and fixed two other problems: it was poking at a workqueue owned by the mmc core; and in one (rare) case would try freeing an IRQ that it didn't allocate. Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Nicolas Ferre <nicolas.ferre@atmel.com> Signed-off-by: Pierre Ossman <drzeus@drzeus.cx> |
||
---|---|---|
.. | ||
at91_mci.c | ||
au1xmmc.c | ||
au1xmmc.h | ||
imxmmc.c | ||
imxmmc.h | ||
Kconfig | ||
Makefile | ||
mmc_spi.c | ||
mmci.c | ||
mmci.h | ||
omap.c | ||
pxamci.c | ||
pxamci.h | ||
ricoh_mmc.c | ||
sdhci.c | ||
sdhci.h | ||
tifm_sd.c | ||
wbsd.c | ||
wbsd.h |