s-paulraj@ti.com wrote:

From: Sandeep Paulraj s-paulraj@ti.com

This patch fixes a compilation warning while building DM355

Signed-off-by: Sandeep Paulraj s-paulraj@ti.com

board/davinci/dm355evm/dm355evm.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/board/davinci/dm355evm/dm355evm.c b/board/davinci/dm355evm/dm355evm.c index 0a44748..5dd3b0d 100644 --- a/board/davinci/dm355evm/dm355evm.c +++ b/board/davinci/dm355evm/dm355evm.c @@ -96,11 +96,11 @@ static void nand_dm355evm_select_chip(struct mtd_info *mtd, int chip) u32 rbase = (u32) this->IO_ADDR_R;

if (chip == 1) {

• __set_bit(14, &wbase);
  

• __set_bit(14, &rbase);
  

• __set_bit(14, (void *)&wbase);
  

• __set_bit(14, (void *)&rbase);
  

  } else {

• __clear_bit(14, &wbase);
  

• __clear_bit(14, &rbase);
  

• __clear_bit(14, (void *)&wbase);
  

• __clear_bit(14, (void *)&rbase);
  

  } this->IO_ADDR_W = (void *)wbase; this->IO_ADDR_R = (void *)rbase;

It looks like __clear_bit is #defined to generic_clear_bit which has this declaration. generic_clear_bit(int nr, volatile unsigned long addr) So maybe changing u32 to volatile unsigned long will work?

Also it looks like dm365evm has the same logic for chip select. At the least, it should also be fixed. Since they look like copies, it would be better if there was a common function.

Tom