[PATCH 1/5] mtd: brcmnand: Fix pointer type-cast in brcmnand_write()
From: computersforpeace@gmail.com (Brian Norris)
Date: 2015-10-12 21:20:21
Also in:
linux-devicetree, lkml
On Fri, Oct 02, 2015 at 11:26:42PM +0530, Anup Patel wrote:
quoted hunk ↗ jump to hunk
We should always type-cast pointer to "long" or "unsigned long" because size of pointer is same as machine word size. This will avoid pointer type-cast issues on both 32bit and 64bit systems. This patch fixes pointer type-cast issue in brcmnand_write() as-per above info. Signed-off-by: Anup Patel <redacted> Reviewed-by: Vikram Prakash <redacted> Reviewed-by: Ray Jui <rjui@broadcom.com> Reviewed-by: Scott Branden <sbranden@broadcom.com> --- drivers/mtd/nand/brcmnand/brcmnand.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)diff --git a/drivers/mtd/nand/brcmnand/brcmnand.c b/drivers/mtd/nand/brcmnand/brcmnand.c index fddb795..4cba03d 100644 --- a/drivers/mtd/nand/brcmnand/brcmnand.c +++ b/drivers/mtd/nand/brcmnand/brcmnand.c@@ -1544,9 +1544,9 @@ static int brcmnand_write(struct mtd_info *mtd, struct nand_chip *chip, dev_dbg(ctrl->dev, "write %llx <- %p\n", (unsigned long long)addr, buf); - if (unlikely((u32)buf & 0x03)) { + if (unlikely((unsigned long)buf & 0x03)) { dev_warn(ctrl->dev, "unaligned buffer: %p\n", buf); - buf = (u32 *)((u32)buf & ~0x03); + buf = (u32 *)((unsigned long)buf & ~0x03); }
Hmm, this block isn't really good. We should probably work to remove this. Not sure why I left that in, actually. At a minimum, it'd probably good to make this noisier, so if we see this in practice, we can work to fix those cases. But anyway, the patch is good (though 'uintptr_t' might be slightly better), so pushed to l2-mtd.git. Brian
brcmnand_wp(mtd, 0); -- 1.9.1