On Tue, Dec 21, 2010 at 1:35 AM, Dirk Behme dirk.behme@googlemail.com wrote:

(Resend with corrected broken example)

On 21.12.2010 08:21, Albert ARIBAUD wrote:

...

Hi Dirk,

Le 21/12/2010 08:11, Dirk Behme a écrit :

...

But the issue with drivers/mtd/nand/omap_gpmc.c (i.e. the additional ldrb    r3, [r3]) is still open? Has anybody tried to replace it with a nop in the binary to be sure this is the root cause?

Can you try and preprocess the C file for both the broken and working cases, then post the preprocessed C extract? Differences at the C level may help understanding differences at the asm level.

gcc version 4.5.1 (Sourcery G++ Lite 2010.09-50)

Work:

static void omap_nand_hwcontrol(struct mtd_info *mtd, int32_t cmd,     uint32_t ctrl) {  register struct nand_chip *this = mtd->priv;  ...  if (cmd != -1)

(*(volatile unsigned char *)(this->IO_ADDR_W) = (cmd)); }

if (cmd != NAND_CMD_NONE)   84:  e3710001        cmn     r1, #1                origwriteb(cmd, this->IO_ADDR_W);   88:  15933004        ldrne   r3, [r3, #4]   8c:  120110ff        andne   r1, r1, #255    ; 0xff   90:  15c31000        strbne  r1, [r3]   94:  e12fff1e        bx      lr        ...

Broken:

static void omap_nand_hwcontrol(struct mtd_info *mtd, int32_t cmd,     uint32_t ctrl) {  register struct nand_chip *this = mtd->priv;

...

if (cmd != -1)   ({ __asm__ __volatile__ ("" : : : "memory"); (*(volatile unsigned char *)(this->IO_ADDR_W) = (cmd)); }); }

if (cmd != NAND_CMD_NONE)   84:  e3710001        cmn     r1, #1   88:  012fff1e        bxeq    lr                writeb(cmd, this->IO_ADDR_W);   8c:  e5933004        ldr     r3, [r3, #4]   90:  e20110ff        and     r1, r1, #255    ; 0xff   94:  e5c31000        strb    r1, [r3]   98:  e5d33000        ldrb    r3, [r3]   9c:  e12fff1e        bx      lr

The issue seems to be the additional 'ldrb      r3, [r3]' added by the compiler in the broken version.

And I at your suggestion tried modifying the binary changing the extra ldrb to a nop and it works.