Re: [PATCH v5 03/10] [media] marvell-ccic: don't generate EOF on parallel bus
From: Jacopo Mondi <jacopo@jmondi.org>
Date: 2019-06-14 10:00:28
Also in:
linux-media, lkml
Hi Lubomir, On Sun, May 05, 2019 at 04:00:24PM +0200, Lubomir Rintel wrote:
The commit 05fed81625bf ("[media] marvell-ccic: add MIPI support for
marvell-ccic driver") that claimed to add CSI2 turned on C0_EOF_VSYNC for
parallel bus without a very good explanation.
That broke camera on OLPC XO-1.75 which precisely uses a sensor on a
parallel bus. Revert that chunk.
Tested on an OLPC XO-1.75.
Fixes: 05fed81625bf755cc67c5864cdfd18b69ea828d1
Use the proper fixes format here
Fixes: 05fed81625bf ("[media] marvell-ccic: add MIPI support for marvell-ccic driver")
I have this simple entry in my git config:
[pretty]
fixes = Fixes: %h (\"%s\")
With
abbrev=12
in the [core] section.
You can now
$git show 05fed81625bf755cc67c5864cdfd18b69ea828d1 --pretty=fixes
quoted hunk ↗ jump to hunk
Signed-off-by: Lubomir Rintel <lkundrak@v3.sk> --- drivers/media/platform/marvell-ccic/mcam-core.c | 6 ------ 1 file changed, 6 deletions(-)diff --git a/drivers/media/platform/marvell-ccic/mcam-core.c b/drivers/media/platform/marvell-ccic/mcam-core.c index d97f39bde9bd6..d24e5b7a3bc52 100644 --- a/drivers/media/platform/marvell-ccic/mcam-core.c +++ b/drivers/media/platform/marvell-ccic/mcam-core.c@@ -792,12 +792,6 @@ static void mcam_ctlr_image(struct mcam_camera *cam) * Make sure it knows we want to use hsync/vsync. */ mcam_reg_write_mask(cam, REG_CTRL0, C0_SIF_HVSYNC, C0_SIFM_MASK); - /* - * This field controls the generation of EOF(DVP only) - */ - if (cam->bus_type != V4L2_MBUS_CSI2_DPHY) - mcam_reg_set_bit(cam, REG_CTRL0, - C0_EOF_VSYNC | C0_VEDGE_CTRL);
This change seems in facts unrelated to the original patch. As you remove all usages of C0_EOF_VSYNC and C0_VEDGE_CTRL you can drop their definition in mcam-core.h. As I've said, the change seems unrelated to CSI-2 support and could probably be salfey dropped, but pay attention, you're also dropping C0_VEDGE_CTRL, which seems to enable VSYNC detection on the signal falling edge. Is this intentional ? Thanks j
} -- 2.21.0
Attachments
- signature.asc [application/pgp-signature] 833 bytes