| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <YcHDQ1hwPLNnH0CN@paasikivi.fi.intel.com>
Date: Tue, 21 Dec 2021 14:06:27 +0200
From: Sakari Ailus <sakari.ailus@...ux.intel.com>
To: hugues.fruchet@...com, Eugen.Hristev@...rochip.com
Cc: prabhakar.mahadev-lad.rj@...renesas.com, jacopo+renesas@...ndi.org,
laurent.pinchart@...asonboard.com, slongerbeam@...il.com,
paul.kocialkowski@...tlin.com, linux-media@...r.kernel.org,
linux-kernel@...r.kernel.org, linux-renesas-soc@...r.kernel.org,
biju.das.jz@...renesas.com, prabhakar.csengg@...il.com,
mchehab@...nel.org
Subject: Re: [PATCH v4 1/6] media: i2c: ov5640: Remain in power down for DVP
mode unless streaming
Hi Hugues, Eugen,
On Tue, Dec 21, 2021 at 07:37:47AM +0000, Eugen.Hristev@...rochip.com wrote:
> On 6/29/21 1:47 PM, Eugen Hristev - M18282 wrote:
> > On 9/9/20 7:16 PM, Hugues FRUCHET wrote:
> >> Hi Prabhakar,
> >>
> >> As discussed separately I would prefer to better understand issue before
> >> going to this patch.
> >> Nevertheless I have some remarks in code in case we'll need it at the end.
> >>
> >> On 9/4/20 10:18 PM, Lad Prabhakar wrote:
> >>> Keep the sensor in software power down mode and wake up only in
> >>> ov5640_set_stream_dvp() callback.
> >>>
> >>> Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@...renesas.com>
> >>> Reviewed-by: Biju Das <biju.das.jz@...renesas.com>
> >>> Tested-by: Jacopo Mondi <jacopo@...ndi.org>
> >>> ---
> >>> drivers/media/i2c/ov5640.c | 19 ++++++++++++++++---
> >>> 1 file changed, 16 insertions(+), 3 deletions(-)
> >>>
> >>> diff --git a/drivers/media/i2c/ov5640.c b/drivers/media/i2c/ov5640.c
> >>> index 2fe4a7ac0592..880fde73a030 100644
> >>> --- a/drivers/media/i2c/ov5640.c
> >>> +++ b/drivers/media/i2c/ov5640.c
> >>> @@ -34,6 +34,8 @@
> >>> #define OV5640_REG_SYS_RESET02 0x3002
> >>> #define OV5640_REG_SYS_CLOCK_ENABLE02 0x3006
> >>> #define OV5640_REG_SYS_CTRL0 0x3008
> >>> +#define OV5640_REG_SYS_CTRL0_SW_PWDN 0x42
> >>> +#define OV5640_REG_SYS_CTRL0_SW_PWUP 0x02
> >>
> >> For the time being this section was only referring to registers
> >> addresses and bit details was explained into a comment right before
> >> affectation, see OV5640_REG_IO_MIPI_CTRL00 for example.
> >>
> >>> #define OV5640_REG_CHIP_ID 0x300a
> >>> #define OV5640_REG_IO_MIPI_CTRL00 0x300e
> >>> #define OV5640_REG_PAD_OUTPUT_ENABLE01 0x3017
> >>> @@ -1120,6 +1122,12 @@ static int ov5640_load_regs(struct ov5640_dev *sensor,
> >>> val = regs->val;
> >>> mask = regs->mask;
> >>>
> >>> + /* remain in power down mode for DVP */
> >>> + if (regs->reg_addr == OV5640_REG_SYS_CTRL0 &&
> >>> + val == OV5640_REG_SYS_CTRL0_SW_PWUP &&
> >>> + sensor->ep.bus_type != V4L2_MBUS_CSI2_DPHY)
> >>> + continue;
> >>> +
> >>
> >> I understand that more or less register OV5640_REG_SYS_CTRL0 (0x3008)
> >> has been partially removed from big init sequence: for power-up part,
> >> but power-dwn remains at very beginning of sequence.
> >> We should completely remove 0x3008 from init sequence, including
> >> power-dwn, and introduce a new function ov5640_sw_powerdown(on/off) that
> >> should be called at the right place instead.
> >>
> >>
> >>> if (mask)
> >>> ret = ov5640_mod_reg(sensor, reg_addr, mask, val);
> >>> else
> >>> @@ -1297,9 +1305,14 @@ static int ov5640_set_stream_dvp(struct ov5640_dev *sensor, bool on)
> >>> * PAD OUTPUT ENABLE 02
> >>> * - [7:2]: D[5:0] output enable
> >>> */
> >>> - return ov5640_write_reg(sensor,
> >>> - OV5640_REG_PAD_OUTPUT_ENABLE02,
> >>> - on ? 0xfc : 0);
> >>> + ret = ov5640_write_reg(sensor, OV5640_REG_PAD_OUTPUT_ENABLE02,
> >>> + on ? 0xfc : 0);
> >>> + if (ret)
> >>> + return ret;
> >>> +
> >>> + return ov5640_write_reg(sensor, OV5640_REG_SYS_CTRL0, on ?
> >>> + OV5640_REG_SYS_CTRL0_SW_PWUP :
> >>> + OV5640_REG_SYS_CTRL0_SW_PWDN);
> >>> }
> >>>
> >>> static int ov5640_set_stream_mipi(struct ov5640_dev *sensor, bool on)
> >>>
> >>
> >>
> >> BR,
> >> Hugues.
> >>
> >
> >
> > Hello everyone,
> >
> > When I updated driver in my tree with this patch, I noticed that my
> > setup using the ATMEL isc platform + ov5640 in parallel bus mode stopped
> > working.
> >
> > It looks like the culprit is this patch.
> > I am not sure which is the best way to fix it.
> > It looks like initially things work fine when probing the driver, but
> > when we want to start streaming at a later point, the register SYS_CTRL0
> > cannot be written anymore.
> > Here is an output of the log:
> >
> > --- Opening /dev/video0...
> > Trying source module v4l2...
> > /dev/video0 opened.
> > No input was specified, using the first.
> > ov5640 1-003c: ov5640_write_reg: error: reg=3008, val=2
> > atmel-sama5d2-isc f0008000.isc: stream on failed in subdev -121
> > Error starting stream.
> > VIDIOC_STREAMON: Remote I/O error
> > Unable to use mmap. Using read instead.
> > Unable to use read.
> >
> > I am using a simple application to read from /dev/video0 (which is
> > registered by the atmel-isc once the sensor probed)
> >
> > I have a feeling that something is wrong related to power, but I cannot
> > tell for sure.
> >
> > Reverting this patch makes it work fine again.
> >
> > Help is appreciated,
> > Thanks,
> > Eugen
> >
>
>
> Gentle ping.
>
> I would like to send a revert patch if nobody cares about this
> driver/patch except me.
Hugues: any thoughts on this?
Without knowing much about the sensor, the stated intent of the patch seems
reasonable. But if it breaks things, then reverting it sounds like a good
idea until a proper fix is available.
--
Kind regards,
Sakari Ailus
Powered by blists - more mailing lists