linux/drivers/mtd/maps, branch v3.18

mtd: physmap_of: Add non-obsolete map_rom probe

2014-09-22T17:13:13Z

Previously, the only way to map a NOR device as a simple ROM was to use the obsolete "direct-mapped" compatible binding (which further requires device_type = "nor" and probe-type = "NOR" properties). This patch adds an "mtd-rom" compatible binding to the "map_rom" probe type. Signed-off-by: Aaron Sierra Signed-off-by: Brian Norris

mtd: physmap_of: Fix ROM support via OF

2014-09-22T17:13:11Z

The "ROM" and unknown probe types within the obsolete "direct-mapped" probe function used the nonexistent "mtd_rom" probe instead of the intended "map_rom". Signed-off-by: Aaron Sierra Signed-off-by: Brian Norris

mtd: gpio_flash: handle case where offset + len exceeds the window size

2014-08-19T18:53:10Z

Fix the bug in handling gpio flash read/write when offset + len from MTD exceeds the window size Signed-off-by: Aaron Wu [Brian: made some commentary edits. Also note that the BUG_ON() was provably false for all non-negative inputs (since x % y <= x), so we dropped it.] Signed-off-by: Brian Norris

mtd: maps: solutionengine: drop excess dependency

2014-08-19T18:53:09Z

Already depends on SOLUTION_ENGINE, so we don't need the SUPERH dependency too. Signed-off-by: Brian Norris

mtd: use NULL instead of 0 for an address

2014-08-19T18:53:05Z

Use NULL instead of 0 when returning an address. This fixes a sparse warning. Signed-off-by: Martin Kepplinger Signed-off-by: Brian Norris

mtd: maps: rbtx4939-flash: delete an unused variable in rbtx4939_flash_remove

2014-07-02T22:21:39Z

Signed-off-by: Atsushi Nemoto Signed-off-by: Brian Norris

mtd: maps: remove check for CONFIG_MTD_SUPERH_RESERVE

2014-05-28T22:12:47Z

Since (a few releases before) v2.6.0 there have been checks for CONFIG_MTD_SUPERH_RESERVE. One check is still present. But a Kconfig symbol MTD_SUPERH_RESERVE has never been added. So a few lines of dead code can be removed. Signed-off-by: Paul Bolle Signed-off-by: Brian Norris

mtd: maps: sc520cdp: fix warnings

2014-05-28T07:05:24Z

On m86k, and maybe a few other architectures, we get this kind of warning, due to misuse of volatile: drivers/mtd/maps/sc520cdp.c: In function 'sc520cdp_setup_par': >> drivers/mtd/maps/sc520cdp.c:223:2: warning: passing argument 1 of 'iounmap' discards 'volatile' qualifier from pointer target type [enabled by default] arch/m68k/include/asm/raw_io.h:22:13: note: expected 'void *' but argument is of type 'volatile long unsigned int *' Rather than annotating the variable declaration, let's just use the proper accessors, which add the 'volatile' qualifier to the operation. Signed-off-by: Brian Norris

mtd: Fix Elan SC520 dependencies

2014-05-12T18:55:53Z

Anyone working with an AMD Elan SC520 development or evaluation board would be building a dedicated kernel for it, so we can make the sc520cdp and netsc520 maps depend on MELAN. SC520_CPUFREQ already depends on MELAN so it makes things more consistent. It also makes kernel configuration for every other x86 user easier. Signed-off-by: Jean Delvare Cc: David Woodhouse Signed-off-by: Brian Norris

mtd: allow CONFIG_MTD_PHYSMAP_OF also for CONFIG_MTD_RAM

2014-03-11T05:42:29Z

Up to now mtd-ram devices described in device trees were only accessible if mtd-flash or mtd-rom were also configured at linux configuration time, because MTD_PHYSMAP_OF was only available if (MTD_CFI || MTD_JEDECPROBE || MTD_ROM). Allow MTD_PHYSMAP_OF selection also when only MTD_RAM is set. Signed-off-by: Philippe De Muyter Signed-off-by: Brian Norris