Message ID | 1405503532-6415-1-git-send-email-l.majewski@samsung.com |
---|---|
State | Rejected |
Delegated to: | Marek Vasut |
Headers | show |
On 07/16/2014 03:38 AM, Lukasz Majewski wrote: > Commit 1151b7ac10b81ecbb has cleaned up read and write operations. > Unfortunately, for correct operation the write for ext4 fs requires > absolute patch. > This patch fixes this case. Shouldn't the user simply supply the absolute patch name as the entity name? That's what I've been doing. This way, it makes files in sub-directories look more natural; you'd pass in e.g. /boot/zImage rather than boot/zImage, which looks like a relative path.
Hi Stephen, > On 07/16/2014 03:38 AM, Lukasz Majewski wrote: > > Commit 1151b7ac10b81ecbb has cleaned up read and write operations. > > Unfortunately, for correct operation the write for ext4 fs requires > > absolute patch. > > This patch fixes this case. > > Shouldn't the user simply supply the absolute patch name as the entity > name? That's what I've been doing. > > This way, it makes files in sub-directories look more natural; you'd > pass in e.g. /boot/zImage rather than boot/zImage, which looks like a > relative path. Hmm.. this is the angle from which I didn't look. Thanks for the tip. I will adjust envs accordingly. Please discard this patch.
On Tuesday, July 22, 2014 at 10:00:05 AM, Lukasz Majewski wrote: > Hi Stephen, > > > On 07/16/2014 03:38 AM, Lukasz Majewski wrote: > > > Commit 1151b7ac10b81ecbb has cleaned up read and write operations. > > > Unfortunately, for correct operation the write for ext4 fs requires > > > absolute patch. > > > This patch fixes this case. > > > > Shouldn't the user simply supply the absolute patch name as the entity > > name? That's what I've been doing. > > > > This way, it makes files in sub-directories look more natural; you'd > > pass in e.g. /boot/zImage rather than boot/zImage, which looks like a > > relative path. > > Hmm.. this is the angle from which I didn't look. > > Thanks for the tip. I will adjust envs accordingly. Please discard this > patch. Done. Best regards, Marek Vasut
diff --git a/drivers/dfu/dfu_mmc.c b/drivers/dfu/dfu_mmc.c index 72fa03e..38aeab0 100644 --- a/drivers/dfu/dfu_mmc.c +++ b/drivers/dfu/dfu_mmc.c @@ -117,6 +117,7 @@ static int mmc_file_op(enum dfu_op op, struct dfu_entity *dfu, { const char *fsname, *opname; char cmd_buf[DFU_CMD_BUF_SIZE]; + char *filename = " %s"; char *str_env; int ret; @@ -153,7 +154,10 @@ static int mmc_file_op(enum dfu_op op, struct dfu_entity *dfu, if (op != DFU_OP_SIZE) sprintf(cmd_buf + strlen(cmd_buf), " 0x%x", (unsigned int)buf); - sprintf(cmd_buf + strlen(cmd_buf), " %s", dfu->name); + if (dfu->layout == DFU_FS_EXT4 && op == DFU_OP_WRITE) + filename = " /%s"; + + sprintf(cmd_buf + strlen(cmd_buf), filename, dfu->name); if (op == DFU_OP_WRITE) sprintf(cmd_buf + strlen(cmd_buf), " %lx", *len);
Commit 1151b7ac10b81ecbb has cleaned up read and write operations. Unfortunately, for correct operation the write for ext4 fs requires absolute patch. This patch fixes this case. Signed-off-by: Lukasz Majewski <l.majewski@samsung.com> --- drivers/dfu/dfu_mmc.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)