Re: Initialize v4l2_requestbuffers struct to avoid invalid mmap

[Ingo Feinerer]

On Thu, May 28, 2020 at 12:23:58PM +0200, Martin Pieuchot wrote:
> On 26/05/20(Tue) 11:30, Ingo Feinerer wrote:
> > video(1) supports reading frames from a webcam via mmap(). To inform the
> > V4L2 device about the number of desired buffers containing the frames to
> > be memory-mapped, a VIDIOC_REQBUFS ioctl call is used.
> > 
> > At the moment the v4l2_requestbuffers struct used for the VIDIOC_REQBUFS 
> > ioctl
> > is only partially initialized which leads to an invalid mmap for my webcam:
> 
> The kernel driver, uvideo(4) or utvfu(4) is responsible for the rest of
> the initialization.

Thanks for the hint. uvideo_reqbufs() is ignoring both struct members
`type` and `memory` so it makes no difference when using video(1) as is.
I wrote a small program that just fetches one frame and can confirm it
works.

However, I have a camera that only supports the MJPEG encoding which
video(1) cannot handle. So I am using libv4l (pkg_add libv4l) which
can convert encodings on the fly:

LD_PRELOAD=/usr/local/lib/v4l2convert.so video

libv4l inspects the `type` and `memory` struct members
(https://git.linuxtv.org/v4l-utils.git/tree/lib/libv4l2/libv4l2.c#n1308)
and fails if `memory != V4L2_MEMORY_MMAP`.

So the real question is if video(1) should "support" libv4l ...

Best regards,
Ingo

Initialize v4l2_requestbuffers struct to avoid invalid mmap

[Ingo Feinerer]

video(1) supports reading frames from a webcam via mmap(). To inform the
V4L2 device about the number of desired buffers containing the frames to
be memory-mapped, a VIDIOC_REQBUFS ioctl call is used.

At the moment the v4l2_requestbuffers struct used for the VIDIOC_REQBUFS ioctl
is only partially initialized which leads to an invalid mmap for my webcam:

video: mmap: Invalid argument

With the following diff applied my camera works (only on EHCI ports with USB
3.0 disabled in BIOS but that is a different story).

OK?

Index: video.c
===
RCS file: /cvs/xenocara/app/video/video.c,v
retrieving revision 1.29
diff -u -p -r1.29 video.c
--- video.c 6 Nov 2019 05:46:51 -   1.29
+++ video.c 26 May 2020 09:14:39 -
@@ -1354,6 +1354,8 @@ mmap_init(struct video *vid)
 
/* request buffers */
rb.count = MMAP_NUM_BUFS;
+   rb.type = V4L2_BUF_TYPE_VIDEO_CAPTURE;
+   rb.memory = V4L2_MEMORY_MMAP;
r = ioctl(vid->dev.fd, VIDIOC_REQBUFS, );
if (r == -1) {
warn("ioctl VIDIOC_REQBUFS");

Re: problems with libm

[Ingo Feinerer]

Moritz Buhl wrote:
> ... I noticed that some floating point operations cause failures of other 
> tests.
> ...
> Many edge cases for complex floating point operations are not covered at all.

Hi,

https://marc.info/?l=openbsd-tech=150737856618497=2 is another example of
an edge case for complex floating point operations.

https://github.com/wch/r-source/blob/trunk/src/main/complex.c#L452-L455 gives
a solution by checking if the imaginary part of the input complex number is
too large (as otherwise sinh() is called which grows exponentially (see e.g.
https://www.wolframalpha.com/input/?i=sinh(x) ) resulting in an overflow.)

Note that the ctan() implementation in R is under GPL, so I am unsure if the
check can be taken as is and committed to OpenBSD.

s_ctanf.c probably needs a similar treatment.

Best regards,
Ingo

Index: s_ctan.c
===
RCS file: /cvs/src/lib/libm/src/s_ctan.c,v
retrieving revision 1.7
diff -u -p -r1.7 s_ctan.c
--- s_ctan.c12 Sep 2016 19:47:02 -  1.7
+++ s_ctan.c11 Jul 2019 12:31:41 -
@@ -135,9 +135,11 @@ double complex
 ctan(double complex z)
 {
double complex w;
-   double d;
+   double d, wy, x, y;
 
-   d = cos (2.0 * creal (z)) + cosh (2.0 * cimag (z));
+   x = 2.0 * creal(z);
+   y = 2.0 * cimag(z);
+   d = cos(x) + cosh(y);
 
if (fabs(d) < 0.25)
d = _ctans (z);
@@ -148,7 +150,12 @@ ctan(double complex z)
return (w);
}
 
-   w = sin (2.0 * creal(z)) / d + (sinh (2.0 * cimag(z)) / d) * I;
+   if (isnan(y) || fabs(y) < 50.0)
+   wy = sinh(y) / d;
+   else
+   wy = (y < 0 ? -1.0 : 1.0);
+
+   w = sin(x) / d + wy * I;
return (w);
 }
 DEF_STD(ctan);

if_umb.c: typos

[Ingo Feinerer]

A few messsage -> message fixes.

Index: if_umb.c
===
RCS file: /cvs/src/sys/dev/usb/if_umb.c,v
retrieving revision 1.21
diff -u -p -r1.21 if_umb.c
--- if_umb.c2 Oct 2018 19:49:10 -   1.21
+++ if_umb.c14 Jan 2019 13:16:50 -
@@ -1197,7 +1197,7 @@ umb_decode_response(struct umb_softc *sc
umb_command_done(sc, response, len);
break;
default:
-   DPRINTF("%s: discard messsage %s\n", DEVNAM(sc),
+   DPRINTF("%s: discard message %s\n", DEVNAM(sc),
umb_request2str(type));
break;
}
@@ -1211,19 +1211,19 @@ umb_handle_indicate_status_msg(struct um
uint32_t cid;
 
if (len < sizeof (*m)) {
-   DPRINTF("%s: discard short %s messsage\n", DEVNAM(sc),
+   DPRINTF("%s: discard short %s message\n", DEVNAM(sc),
umb_request2str(letoh32(m->hdr.type)));
return;
}
if (memcmp(m->devid, umb_uuid_basic_connect, sizeof (m->devid))) {
-   DPRINTF("%s: discard %s messsage for other UUID '%s'\n",
+   DPRINTF("%s: discard %s message for other UUID '%s'\n",
DEVNAM(sc), umb_request2str(letoh32(m->hdr.type)),
umb_uuid2str(m->devid));
return;
}
infolen = letoh32(m->infolen);
if (len < sizeof (*m) + infolen) {
-   DPRINTF("%s: discard truncated %s messsage (want %d, got %d)\n",
+   DPRINTF("%s: discard truncated %s message (want %d, got %d)\n",
DEVNAM(sc), umb_request2str(letoh32(m->hdr.type)),
(int)sizeof (*m) + infolen, len);
return;
@@ -2333,7 +2333,7 @@ umb_command_done(struct umb_softc *sc, v
int  qmimsg = 0;
 
if (len < sizeof (*cmd)) {
-   DPRINTF("%s: discard short %s messsage\n", DEVNAM(sc),
+   DPRINTF("%s: discard short %s message\n", DEVNAM(sc),
umb_request2str(letoh32(cmd->hdr.type)));
return;
}
@@ -2341,7 +2341,7 @@ umb_command_done(struct umb_softc *sc, v
if (memcmp(cmd->devid, umb_uuid_basic_connect, sizeof (cmd->devid))) {
if (memcmp(cmd->devid, umb_uuid_qmi_mbim,
sizeof (cmd->devid))) {
-   DPRINTF("%s: discard %s messsage for other UUID '%s'\n",
+   DPRINTF("%s: discard %s message for other UUID '%s'\n",
DEVNAM(sc), umb_request2str(letoh32(cmd->hdr.type)),
umb_uuid2str(cmd->devid));
return;
@@ -2370,7 +2370,7 @@ umb_command_done(struct umb_softc *sc, v
 
infolen = letoh32(cmd->infolen);
if (len < sizeof (*cmd) + infolen) {
-   DPRINTF("%s: discard truncated %s messsage (want %d, got %d)\n",
+   DPRINTF("%s: discard truncated %s message (want %d, got %d)\n",
DEVNAM(sc), umb_cid2str(cid),
(int)sizeof (*cmd) + infolen, len);
return;

ctan(3) NaN

[Ingo Feinerer]

Hi,

I think I found a bug/inconsistency/regression in ctan(3) introduced
somewhere after 6.1 release. Originally observed in math/R
(https://marc.info/?l=openbsd-ports=150728711530471=2) the behavior can be
reproduced with following C program:

--8<---
#include 
#include 
#include 

int
main(int argc, char *argv[])
{
// http://en.cppreference.com/w/c/numeric/complex/ctan
double complex z = ctan(1);  // behaves like real tangent along the 
real line
printf("tan(1+0i) = %f%+fi ( tan(1)=%f)\n", creal(z), cimag(z), tan(1));

// http://en.cppreference.com/w/c/numeric/complex/ctan
double complex z2 = ctan(I); // behaves like tanh along the imaginary 
line
printf("tan(0+1i) = %f%+fi (tanh(1)=%f)\n", creal(z2), cimag(z2), 
tanh(1));

double complex z3 = ctan(1+1000I);
printf("tan(1+1000i) = %f%+fi\n", creal(z3), cimag(z3));
}
--8<---

cc ctan.c -lm
./a.out

The correct solution for the z3 example is "0+1i".

On a -current amd64 system (both clang and gcc give identical results):

$ ./a.out
tan(1+0i) = 1.557408+0.00i ( tan(1)=1.557408)
tan(0+1i) = 0.00+0.761594i (tanh(1)=0.761594)
tan(1+1000i) = -nan-nani

On 6.1 release amd64 with gcc 4.2.1:

$ ./a.out
tan(1+0i) = 1.557408+0.00i ( tan(1)=1.557408)
tan(0+1i) = 0.00+0.761594i (tanh(1)=0.761594)
tan(1+1000i) = 0.00-nani

This result seems to be "sufficiently different from wrong" that R is somehow
able to return the correct solution on 6.1 release:

R> (z <- tan(1+1000i))
[1] 0+1i

On a Debian 9.1 Linux with gcc 6.3.0:

$ ./a.out
tan(1+0i) = 1.557408+0.00i ( tan(1)=1.557408)
tan(0+1i) = 0.00+0.761594i (tanh(1)=0.761594)
tan(1+1000i) = 0.00+1.00i

Ideas, pointers and solutions are highly appreciated.

Best regards,
Ingo

Re: Sync sys/videoio.h with recent Linux kernel

[Ingo Feinerer]

On Sun, Jan 29, 2017 at 11:23:06AM +0100, Martin Pieuchot wrote:
> On 28/01/17(Sat) 11:55, Ingo Feinerer wrote:
> > in a previous commit
> > (http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys/sys/videoio.h?rev=1.12=text/x-cvsweb-markup)
> > we updated our V4L2 API. However, one file (v4l2-common.h) could not be
> > inlined as it was only GPL2 licensed.
> > 
> > After discussion with the authors they agreed to dual license (GPL + BSD) 
> > this
> > file as well
> > (https://git.linuxtv.org/media_tree.git/commit/?id=80100fd9ebb9f2414892a1178d26a4253e6c0bcf)
> > which allows us to fully comply with the V4L2 API. The attached diff
> > implements this and syncs with videodev2.h from the Linux kernel version
> > 4.10-rc5 at the same time
> > (https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/plain/include/uapi/linux/videodev2.h?id=refs/tags/v4.10-rc5).
> > 
> > (The sync with a recent Linux videodev2.h is a prerequisite for updating the
> > port multimedia/libv4l to 1.12.2.)
> > 
> > I compiled a current kernel on amd64. My camera still works fine (tested 
> > with
> > luvcview and baresip in combination with libv4l 1.12.2).
> > 
> > OK?
> 
> This needs to go in a bulk since most users of this API are ports.

May I ask for OKs after naddy@'s successful amd64 bulk build?

Best regards,
Ingo

Sync sys/videoio.h with recent Linux kernel

[Ingo Feinerer]

Hi,

in a previous commit
(http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/sys/sys/videoio.h?rev=1.12=text/x-cvsweb-markup)
we updated our V4L2 API. However, one file (v4l2-common.h) could not be
inlined as it was only GPL2 licensed.

After discussion with the authors they agreed to dual license (GPL + BSD) this
file as well
(https://git.linuxtv.org/media_tree.git/commit/?id=80100fd9ebb9f2414892a1178d26a4253e6c0bcf)
which allows us to fully comply with the V4L2 API. The attached diff
implements this and syncs with videodev2.h from the Linux kernel version
4.10-rc5 at the same time
(https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/plain/include/uapi/linux/videodev2.h?id=refs/tags/v4.10-rc5).

(The sync with a recent Linux videodev2.h is a prerequisite for updating the
port multimedia/libv4l to 1.12.2.)

I compiled a current kernel on amd64. My camera still works fine (tested with
luvcview and baresip in combination with libv4l 1.12.2).

OK?

Best regards,
Ingo

Index: sys/sys/videoio.h
===
RCS file: /cvs/src/sys/sys/videoio.h,v
retrieving revision 1.14
diff -u -p -r1.14 videoio.h
--- sys/sys/videoio.h   27 Jan 2016 13:07:09 -  1.14
+++ sys/sys/videoio.h   28 Jan 2017 10:45:55 -
@@ -3,6 +3,8 @@
  *  Video for Linux Two header file
  *
  *  Copyright (C) 1999-2012 the contributors
+ *  Copyright (C) 2012 Nokia Corporation
+ *  Contact: Sakari Ailus 
  *
  *  This program is free software; you can redistribute it and/or modify
  *  it under the terms of the GNU General Public License as published by
@@ -47,7 +49,7 @@
  * All kernel-specific stuff were moved to media/v4l2-dev.h, so
  * no #if __KERNEL tests are allowed here
  *
- * See http://linuxtv.org for more info
+ * See https://linuxtv.org for more info
  *
  * Author: Bill Dirks 
  * Justin Schoeman
@@ -60,8 +62,84 @@
 #include 
 #include 
 #include 
-/* Ignore #include  as it is GPL2.
- * Inline #include 
+#ifndef __u8
+typedef u_int8_t __u8;
+#endif
+#ifndef __u16
+typedef u_int16_t __u16;
+#endif
+#ifndef __u32
+typedef u_int32_t __u32;
+#endif
+#ifndef __u64
+typedef u_int64_t __u64;
+#endif
+#ifndef __s32
+typedef int32_t __s32;
+#endif
+#ifndef __s64
+typedef int64_t __s64;
+#endif
+#ifndef __le32
+typedef __u32 __le32;
+#endif
+/* Inline #include 
+ * (v4l2-common.h was split off from videodev2.h and has the same BSD license.)
+ *
+ * Begin of v4l2-common.h
+ */
+/*
+ *
+ * Selection interface definitions
+ *
+ */
+
+/* Current cropping area */
+#define V4L2_SEL_TGT_CROP  0x
+/* Default cropping area */
+#define V4L2_SEL_TGT_CROP_DEFAULT  0x0001
+/* Cropping bounds */
+#define V4L2_SEL_TGT_CROP_BOUNDS   0x0002
+/* Native frame size */
+#define V4L2_SEL_TGT_NATIVE_SIZE   0x0003
+/* Current composing area */
+#define V4L2_SEL_TGT_COMPOSE   0x0100
+/* Default composing area */
+#define V4L2_SEL_TGT_COMPOSE_DEFAULT   0x0101
+/* Composing bounds */
+#define V4L2_SEL_TGT_COMPOSE_BOUNDS0x0102
+/* Current composing area plus all padding pixels */
+#define V4L2_SEL_TGT_COMPOSE_PADDED0x0103
+
+/* Backward compatibility target definitions --- to be removed. */
+#define V4L2_SEL_TGT_CROP_ACTIVE   V4L2_SEL_TGT_CROP
+#define V4L2_SEL_TGT_COMPOSE_ACTIVEV4L2_SEL_TGT_COMPOSE
+#define V4L2_SUBDEV_SEL_TGT_CROP_ACTUALV4L2_SEL_TGT_CROP
+#define V4L2_SUBDEV_SEL_TGT_COMPOSE_ACTUAL V4L2_SEL_TGT_COMPOSE
+#define V4L2_SUBDEV_SEL_TGT_CROP_BOUNDSV4L2_SEL_TGT_CROP_BOUNDS
+#define V4L2_SUBDEV_SEL_TGT_COMPOSE_BOUNDS V4L2_SEL_TGT_COMPOSE_BOUNDS
+
+/* Selection flags */
+#define V4L2_SEL_FLAG_GE   (1 << 0)
+#define V4L2_SEL_FLAG_LE   (1 << 1)
+#define V4L2_SEL_FLAG_KEEP_CONFIG  (1 << 2)
+
+/* Backward compatibility flag definitions --- to be removed. */
+#define V4L2_SUBDEV_SEL_FLAG_SIZE_GE   V4L2_SEL_FLAG_GE
+#define V4L2_SUBDEV_SEL_FLAG_SIZE_LE   V4L2_SEL_FLAG_LE
+#define V4L2_SUBDEV_SEL_FLAG_KEEP_CONFIG V4L2_SEL_FLAG_KEEP_CONFIG
+
+struct v4l2_edid {
+   __u32 pad;
+   __u32 start_block;
+   __u32 blocks;
+   __u32 reserved[5];
+   __u8  *edid;
+};
+/*
+ * End of v4l2-common.h
+ */
+/* Inline #include 
  * (v4l2-controls.h was split off from videodev2.h and has the same BSD 
license.)
  *
  * Begin of v4l2-controls.h
@@ -175,8 +253,10 @@ enum v4l2_colorfx {
  * We reserve 16 controls for this driver. */
 #define V4L2_CID_USER_S2255_BASE   (V4L2_CID_USER_BASE + 0x1030)
 
-/* The base for the si476x driver controls. See include/media/si476x.h for the 
list
- * of controls. Total of 16 controls is reserved for this driver */
+/*
+ * The base for the si476x driver controls. See include/media/drv-intf/si476x.h
+ * for the list of controls. Total of 16 controls is reserved for this driver
+ */
 #define V4L2_CID_USER_SI476X_BASE  (V4L2_CID_USER_BASE + 0x1040)
 
 /* The base for the TI VPE driver 

Re: List Ericsson N5321gw as supported umb(4) device

[Ingo Feinerer]

On Thu, Nov 24, 2016 at 07:01:52PM +0100, Otto Moerbeek wrote:
> On Wed, Nov 23, 2016 at 08:56:53PM +0100, Ingo Feinerer wrote:
> > the "Lenovo N5321gw" is working for me with umb(4)
> > (https://marc.info/?l=openbsd-misc=147992993614369=2). So mention it
> > as a supported device in umb.4?
> > 
> > (Lenovo seems to call it Ericsson N5321gw on its support sites, so I'll
> > use this wording.)
> > 
> > OK?
> 
> This is a different card than Lenovo H5321gw, right?  I have that one,
> but there are some problems with which make it work non-optimal.
> These problems are being ironed out as we speak.

I do not know. The Lenovo support site is quite confusing for me:
http://support.lenovo.com/downloads/ds039685

Lenovo lists a single firmware for H5321gw/C5621gw/N5321gw but on the other
side different names are explicitly used:

$ usbdevs -v
port 4 addr 2: high speed, self powered, config 1, N5321 gw(0x193e), 
Lenovo(0x0bdb), rev 0.00

Best regards,
Ingo

List Ericsson N5321gw as supported umb(4) device

[Ingo Feinerer]

Hi,

the "Lenovo N5321gw" is working for me with umb(4)
(https://marc.info/?l=openbsd-misc=147992993614369=2). So mention it
as a supported device in umb.4?

(Lenovo seems to call it Ericsson N5321gw on its support sites, so I'll
use this wording.)

OK?

Best regards,
Ingo

Index: umb.4
===
RCS file: /cvs/src/share/man/man4/umb.4,v
retrieving revision 1.4
diff -u -p -r1.4 umb.4
--- umb.4   22 Aug 2016 06:45:12 -  1.4
+++ umb.4   23 Nov 2016 19:27:23 -
@@ -44,6 +44,7 @@ PIN again even if the system was reboote
 The following devices should work:
 .Pp
 .Bl -tag -width Ds -offset indent -compact
+.It Ericsson N5321gw
 .It Medion Mobile S4222 (MediaTek OEM)
 .It Sierra Wireless EM8805
 .It Sierra Wireless MC8305

Medion Mobile S4222 works with umb(4)

[Ingo Feinerer]

Hi,

just a short report that the "Medion Mobile S4222" UMTS stick (attached via
USB) works fine with the new umb(4):

umsm0 at uhub0 port 2 configuration 1 interface 0 "MediaTek Inc Product" rev 
2.00/3.00 addr 10
umsm0 detached
umb1 at uhub0 port 2 configuration 1 interface 0 "MediaTek Inc Product" rev 
2.00/3.00 addr 10
umsm0 at uhub0 port 2 configuration 1 interface 2 "MediaTek Inc Product" rev 
2.00/3.00 addr 10
ucom3 at umsm0
umsm1 at uhub0 port 2 configuration 1 interface 3 "MediaTek Inc Product" rev 
2.00/3.00 addr 10
ucom4 at umsm1
umsm2 at uhub0 port 2 configuration 1 interface 4 "MediaTek Inc Product" rev 
2.00/3.00 addr 10
ucom5 at umsm2
umsm3 at uhub0 port 2 configuration 1 interface 5 "MediaTek Inc Product" rev 
2.00/3.00 addr 10
ucom6 at umsm3

Is it worth documenting this (in analogy to umsm(4))?

(Btw. mdoc(7) states for "Tn": "do not use this in new manuals". However, for
consistency with the existing entries I kept it.)

Best regards,
Ingo

Index: umb.4
===
RCS file: /cvs/src/share/man/man4/umb.4,v
retrieving revision 1.2
diff -u -p -r1.2 umb.4
--- umb.4   20 Jun 2016 09:03:08 -  1.2
+++ umb.4   24 Jun 2016 12:42:30 -
@@ -44,6 +44,7 @@ PIN again even if the system was reboote
 The following devices should work:
 .Pp
 .Bl -tag -width Ds -offset indent -compact
+.It Tn Medion Mobile S4222 (MediaTek OEM)
 .It Tn Sierra Wireless EM8805
 .It Tn Sierra Wireless MC8305
 .El

Adapt video.4 for recent videoio.h change

[Ingo Feinerer]

Hi,

video.4 manual page diff to cope with recent videoio.h change. OK?

Best regards,
Ingo

Index: video.4
===
RCS file: /cvs/src/share/man/man4/video.4,v
retrieving revision 1.12
diff -u -p -r1.12 video.4
--- video.4 18 Oct 2014 08:01:34 -  1.12
+++ video.4 1 Jan 2016 16:29:35 -
@@ -54,31 +54,34 @@ struct v4l2_capability {
u_int8_tbus_info[32];
u_int32_t   version;
u_int32_t   capabilities;
-   u_int32_t   reserved[4];
+   u_int32_t   device_caps;
+   u_int32_t   reserved[3];
 };
 .Ed
 .It Dv VIDIOC_ENUM_FMT Fa "struct v4l2_fmtdesc *"
 Enumerate image formats.
 .Bd -literal
 struct v4l2_fmtdesc {
-   u_int32_t   index;
-   enum v4l2_buf_type  type;
-   u_int32_t   flags;
-   u_int8_tdescription[32];
-   u_int32_t   pixelformat;
-   u_int32_t   reserved[4];
+   u_int32_t   index;
+   u_int32_t   type;
+   u_int32_t   flags;
+   u_int8_tdescription[32];
+   u_int32_t   pixelformat;
+   u_int32_t   reserved[4];
 };
 .Ed
 .It Dv VIDIOC_S_FMT Fa "struct v4l2_format *"
 Set the data format.
 .Bd -literal
 struct v4l2_format {
-   enum v4l2_buf_type  type;
+   u_int32_t   type;
union {
struct v4l2_pix_format  pix;
+   struct v4l2_pix_format_mplane   pix_mp;
struct v4l2_window  win;
struct v4l2_vbi_format  vbi;
struct v4l2_sliced_vbi_format   sliced;
+   struct v4l2_sdr_format  sdr;
u_int8_traw_data[200];
 } fmt;
 };
@@ -99,7 +102,8 @@ struct v4l2_input {
u_int32_t   tuner;
v4l2_std_id std;
u_int32_t   status;
-   u_int32_t   reserved[32];
+   u_int32_t   capabilities;
+   u_int32_t   reserved[3];
 };
 .Ed
 .It Dv VIDIOC_G_INPUT Fa "int *"
@@ -110,10 +114,10 @@ Select the current video input.
 Initiate memory mapping or user pointer I/O.
 .Bd -literal
 struct v4l2_requestbuffers {
-   u_int32_t   count;
-   enum v4l2_buf_type  type;
-   enum v4l2_memorymemory;
-   u_int32_t   reserved[2];
+   u_int32_t   count;
+   u_int32_t   type;
+   u_int32_t   memory;
+   u_int32_t   reserved[2];
 };
 .Ed
 .It Dv VIDIOC_QUERYBUF Fa "struct v4l2_buffer *"
@@ -121,20 +125,22 @@ Query the status of a buffer.
 .Bd -literal
 struct v4l2_buffer {
u_int32_t   index;
-   enum v4l2_buf_type  type;
+   u_int32_t   type;
u_int32_t   bytesused;
u_int32_t   flags;
-   enum v4l2_field field;
+   u_int32_t   field;
struct timeval  timestamp;
struct v4l2_timecodetimecode;
u_int32_t   sequence;
-   enum v4l2_memorymemory;
+   u_int32_t   memory;
union {
-   u_int32_t   offset;
-   unsigned long   userptr;
+   u_int32_t   offset;
+   unsigned long   userptr;
+   struct v4l2_plane   *planes;
+   int32_t fd;
} m;
u_int32_t   length;
-   u_int32_t   input;
+   u_int32_t   reserved2;
u_int32_t   reserved;
 };
 .Ed
@@ -169,7 +175,7 @@ struct v4l2_frmivalemun {
union {
struct v4l2_fract   discrete;
struct v4l2_frmival_stepwisestepwise;
-   } un;
+   };
u_int32_t   reserved[2];
 };
 
@@ -183,7 +189,7 @@ struct v4l2_frmival_stepwise {
 Set streaming parameters.
 .Bd -literal
 struct v4l2_streamparm {
-   enum v4l2_buf_type  type;
+   u_int32_t   type;
union {
struct v4l2_captureparm capture;
struct v4l2_outputparm  output;
@@ -218,15 +224,15 @@ Same structures as for
 Enumerate control items.
 .Bd -literal
 struct v4l2_queryctrl {
-   u_int32_t   id;
-   enum v4l2_ctrl_type type;
-   u_int8_tname[32];
-   int32_t minimum;
-   int32_t maximum;
-   int32_t step;
-   int32_t default_value;
-   u_int32_t   flags;
-   u_int32_t   reserved[2];
+   u_int32_t   id;
+   u_int32_t   type;
+   u_int8_tname[32];
+   int32_t minimum;
+   int32_t maximum;
+   int32_t step;
+   int32_t default_value;
+   u_int32_t   flags;
+   u_int32_t   reserved[2];
 };
 .Ed
 .El
@@ -242,6 

Update V4L2 API (sys/videoio.h)

[Ingo Feinerer]

Hi,

I would like to update our V4L2 (Video for Linux Two) API.

The plan is to sync sys/videoio.h with a recent version of
linux/videodev2.h (which has a BSD license statement in it) from the
Linux kernel sources. However, videodev2.h now includes two additional
headers:

1) linux/v4l2-common.h: GPL2 licensed
2) linux/v4l2-controls.h: has the same BSD license statement as
   videodev2.h as it was split off from it.

The proposed diff completely ignores v4l2-common.h. That should work out
as it is not designed to be included directly, contains only new
#defines and structs (compared to what we have now in videoio.h), and is
very short. As a consequence I had to comment out two lines in videoio.h
which are not present in our current version anyway.

As there is now just v4l2-controls.h left I propose to merge this file
directly back into videoio.h. The alternative would be to have it as a
separate file sys/v4l2-controls.h. In Linux it is not designed to be
included on its own (but just by videodev2.h).

This proposal is a revised version of
https://marc.info/?l=openbsd-cvs=145019666410552=2 (which was
reverted because of GPL2 considerations) which worked out in a ports
bulk build by ajacoutot@. I am pretty confident that skipping
v4l2-common.h does not hurt us, so no new bulk build should be strictly
necessary.

The original version had two OKs but I would like to restart the process
because of the restructuring. So I am looking for feedback and OKs.

(After that I'd like to address whitespace and formatting in videoio.h
to ease further syncs. Moreover, the man page video.4 will be updated as
well.)

Best regards,
Ingo

Index: videoio.h
===
RCS file: /cvs/src/sys/sys/videoio.h,v
retrieving revision 1.11
diff -u -p -r1.11 videoio.h
--- videoio.h   15 Dec 2015 17:01:44 -  1.11
+++ videoio.h   20 Dec 2015 08:20:01 -
@@ -2,7 +2,7 @@
 /*
  *  Video for Linux Two header file
  *
- *  Copyright (C) 1999-2007 the contributors
+ *  Copyright (C) 1999-2012 the contributors
  *
  *  This program is free software; you can redistribute it and/or modify
  *  it under the terms of the GNU General Public License as published by
@@ -59,6 +59,926 @@
 #define _SYS_VIDEOIO_H_
 #include 
 #include 
+/* Ignore #include  as it is GPL2.
+ * Inline #include 
+ * (v4l2-controls.h was split off from videodev2.h and has the same BSD 
license.)
+ *
+ * Begin of v4l2-controls.h
+ */
+/* Control classes */
+#define V4L2_CTRL_CLASS_USER   0x0098  /* Old-style 'user' 
controls */
+#define V4L2_CTRL_CLASS_MPEG   0x0099  /* MPEG-compression 
controls */
+#define V4L2_CTRL_CLASS_CAMERA 0x009a  /* Camera class 
controls */
+#define V4L2_CTRL_CLASS_FM_TX  0x009b  /* FM Modulator 
controls */
+#define V4L2_CTRL_CLASS_FLASH  0x009c  /* Camera flash 
controls */
+#define V4L2_CTRL_CLASS_JPEG   0x009d  /* JPEG-compression 
controls */
+#define V4L2_CTRL_CLASS_IMAGE_SOURCE   0x009e  /* Image source 
controls */
+#define V4L2_CTRL_CLASS_IMAGE_PROC 0x009f  /* Image processing 
controls */
+#define V4L2_CTRL_CLASS_DV 0x00a0  /* Digital Video 
controls */
+#define V4L2_CTRL_CLASS_FM_RX  0x00a1  /* FM Receiver controls 
*/
+#define V4L2_CTRL_CLASS_RF_TUNER   0x00a2  /* RF tuner controls */
+#define V4L2_CTRL_CLASS_DETECT 0x00a3  /* Detection controls */
+
+/* User-class control IDs */
+
+#define V4L2_CID_BASE  (V4L2_CTRL_CLASS_USER | 0x900)
+#define V4L2_CID_USER_BASE V4L2_CID_BASE
+#define V4L2_CID_USER_CLASS(V4L2_CTRL_CLASS_USER | 1)
+#define V4L2_CID_BRIGHTNESS(V4L2_CID_BASE+0)
+#define V4L2_CID_CONTRAST  (V4L2_CID_BASE+1)
+#define V4L2_CID_SATURATION(V4L2_CID_BASE+2)
+#define V4L2_CID_HUE   (V4L2_CID_BASE+3)
+#define V4L2_CID_AUDIO_VOLUME  (V4L2_CID_BASE+5)
+#define V4L2_CID_AUDIO_BALANCE (V4L2_CID_BASE+6)
+#define V4L2_CID_AUDIO_BASS(V4L2_CID_BASE+7)
+#define V4L2_CID_AUDIO_TREBLE  (V4L2_CID_BASE+8)
+#define V4L2_CID_AUDIO_MUTE(V4L2_CID_BASE+9)
+#define V4L2_CID_AUDIO_LOUDNESS(V4L2_CID_BASE+10)
+#define V4L2_CID_BLACK_LEVEL   (V4L2_CID_BASE+11) /* Deprecated */
+#define V4L2_CID_AUTO_WHITE_BALANCE(V4L2_CID_BASE+12)
+#define V4L2_CID_DO_WHITE_BALANCE  (V4L2_CID_BASE+13)
+#define V4L2_CID_RED_BALANCE   (V4L2_CID_BASE+14)
+#define V4L2_CID_BLUE_BALANCE  (V4L2_CID_BASE+15)
+#define V4L2_CID_GAMMA (V4L2_CID_BASE+16)
+#define V4L2_CID_WHITENESS (V4L2_CID_GAMMA) /* Deprecated */
+#define V4L2_CID_EXPOSURE  (V4L2_CID_BASE+17)
+#define V4L2_CID_AUTOGAIN  (V4L2_CID_BASE+18)
+#define V4L2_CID_GAIN  (V4L2_CID_BASE+19)
+#define V4L2_CID_HFLIP 

Add V4L2_FMT_FLAG_EMULATED to videoio.h

[Ingo Feinerer]

I would like to add V4L2_FMT_FLAG_EMULATED to sys/videoio.h as libv4l
(https://github.com/jasperla/openbsd-wip/tree/master/multimedia/libv4l)
returns the flag for emulated (i.e., not natively supported by your
webcam) video formats. In turn some userspace applications (like the SIP
user-agent baresip) look for this flag as well.

OK?

Best regards,
Ingo

Index: videoio.h
===
RCS file: /cvs/src/sys/sys/videoio.h,v
retrieving revision 1.8
diff -u -p -r1.8 videoio.h
--- videoio.h   23 Oct 2014 07:34:48 -  1.8
+++ videoio.h   20 Nov 2015 19:02:10 -
@@ -360,6 +360,7 @@ struct v4l2_fmtdesc {
 };
 
 #define V4L2_FMT_FLAG_COMPRESSED 0x0001
+#define V4L2_FMT_FLAG_EMULATED   0x0002
 
 /*
  *  F R A M E   S I Z E   E N U M E R A T I O N

cdce(4) and MBIM

[Ingo Feinerer]

Has anyone experiences with USB devices (e.g. UMTS modem) offering
Mobile Broadband Interface Model (MBIM) interfaces?

According to
www.usb.org/developers/docs/devclass_docs/MBIM10Errata1_073013.zip MBIM
is comparable to NCM (The largest deviation from NCM 1.0 is that
devices transfer raw IP packets instead of packets with 802.3 headers.)
which is supported by cdce(4).

By hardcoding vendor and device ID in sys/dev/usb/if_cdce.c I managed to
have a modem with MBIM interface attach via cdce:

cdce0 at uhub0 port 4 configuration 1 interface 0 MediaTek Inc Product rev 
2.00/3.00 addr 2
cdce0: address XX:XX:XX:XX:XX:XX

ifconfig appears to work as well

cdce0: flags=8802BROADCAST,SIMPLEX,MULTICAST mtu 1500
lladdr XX:XX:XX:XX:XX:XX
priority: 0

but dhclient gets no response. /var/log/messages contains

cdce0: watchdog timeout
cdce0: usb error on tx: TIMEOUT

Any ideas?

Best regards,
Ingo

Add Medion Mobile S4222 to umsm(4)

[Ingo Feinerer]

A man page diff corresponding to the recent umsm commit by Martin
Pieuchot.

Best regards,
Ingo

Index: share/man/man4/umsm.4
===
RCS file: /cvs/src/share/man/man4/umsm.4,v
retrieving revision 1.89
diff -u -p -r1.89 umsm.4
--- share/man/man4/umsm.4   19 Mar 2014 22:56:44 -  1.89
+++ share/man/man4/umsm.4   6 Jan 2015 18:20:59 -
@@ -81,6 +81,7 @@ driver:
 .It Li IIJMobile 120FU (ZTE OEM) Ta USB
 .It Li Kyocera KPC650 Ta CardBus
 .It Li Medion Mobile S4012 (Huawei E1550 OEM) Ta USB
+.It Li Medion Mobile S4222 (MediaTek OEM) Ta USB
 .It Li Novatel Wireless ES620 Ta USB
 .It Li Novatel Wireless Ovation U727 Ta USB
 .It Li Novatel Wireless U760 Ta USB
@@ -158,6 +159,8 @@ but only the first port can be used to m
 the second one is for management.
 The Option GlobeTrotter HSDPA/HSUPA modems have three serial ports,
 but only the last port can be used to make PPP connections.
+The Medion Mobile S4222 has four serial ports, but only the first
+port can be used to make connections.
 .Pp
 Some modems require multi-link PPP operation.
 For example, the Huawei E1550 has four serial ports.

MediaTek Inc. UMTS stick

[Ingo Feinerer]

Please find attached at the end of this mail a diff for the MediaTek
Inc. DC_4COM (that is the product name that the Linux option driver
uses:
https://raw.githubusercontent.com/torvalds/linux/master/drivers/usb/serial/option.c)
UMTS USB stick (= MEDION S4222). It solves the problems described at
http://article.gmane.org/gmane.os.openbsd.misc/219029 . This gives me

umsm0 at uhub7 port 1 configuration 1 interface 0 MediaTek Inc Product rev 
2.00/3.00 addr 4
umsm0 detached
umsm0 at uhub7 port 1 configuration 1 interface 2 MediaTek Inc Product rev 
2.00/3.00 addr 4
ucom0 at umsm0
umsm1 at uhub7 port 1 configuration 1 interface 3 MediaTek Inc Product rev 
2.00/3.00 addr 4
ucom1 at umsm1
umsm2 at uhub7 port 1 configuration 1 interface 4 MediaTek Inc Product rev 
2.00/3.00 addr 4
ucom2 at umsm2
umsm3 at uhub7 port 1 configuration 1 interface 5 MediaTek Inc Product rev 
2.00/3.00 addr 4
ucom3 at umsm3
umass0 at uhub7 port 1 configuration 1 interface 6 MediaTek Inc Product rev 
2.00/3.00 addr 4
umass0: using SCSI over Bulk-Only
scsibus4 at umass0: 2 targets, initiator 0
sd1 at scsibus4 targ 1 lun 0: MEDIATEK, FLASH DISK, 6225 SCSI0 0/direct 
removable
ugen0 at uhub7 port 1 configuration 1 MediaTek Inc Product rev 2.00/3.00 addr 
4

and ucom0 (cuaU0) now works as expected (successfully tested with cu and pppd).

Please note that I had to ignore the first two interfaces when umsm
attaches (otherwise ucom[0-3] do not work although they attach).

Best regards,
Ingo

Index: sys/dev/usb/usbdevs
===
RCS file: /cvs/src/sys/dev/usb/usbdevs,v
retrieving revision 1.641
diff -u -p -r1.641 usbdevs
--- sys/dev/usb/usbdevs 4 Dec 2014 10:41:42 -   1.641
+++ sys/dev/usb/usbdevs 5 Jan 2015 11:20:33 -
@@ -452,6 +452,7 @@ vendor CTC  0x0e5e  CONWISE Technology
 vendor HAWKING 0x0e66  Hawking
 vendor FOSSIL  0x0e67  Fossil
 vendor GMATE   0x0e7e  G.Mate
+vendor MEDIATEK0x0e8d  MediaTek Inc.
 vendor OTI 0x0ea0  Ours Technology
 vendor PILOTECH0x0eaf  Pilotech
 vendor NOVATECH0x0eb0  Nova Tech
@@ -2824,6 +2825,10 @@ product MCT ML_4500  0x0302  ML-4500
 
 /* MDS products */
 product MDS ISDBT  0x0001  MDS ISDB-T tuner
+
+/* MediaTek Inc. products */
+product MEDIATEK UMASS_INSTALLER   0x0002  USB installer
+product MEDIATEK DC_4COM   0x00a5  UMTS USB modem
 
 /* MEI products */
 product MEI CASHFLOW_SC0x1100  Cashflow-SC
Index: sys/dev/usb/umsm.c
===
RCS file: /cvs/src/sys/dev/usb/umsm.c,v
retrieving revision 1.100
diff -u -p -r1.100 umsm.c
--- sys/dev/usb/umsm.c  12 Jul 2014 21:24:33 -  1.100
+++ sys/dev/usb/umsm.c  5 Jan 2015 11:20:33 -
@@ -109,8 +109,9 @@ struct umsm_type {
 #defineDEV_UMASS5  0x0100
 #defineDEV_UMASS6  0x0200
 #defineDEV_UMASS7  0x0400
+#defineDEV_UMASS8  0x1000
 #define DEV_UMASS  (DEV_UMASS1 | DEV_UMASS2 | DEV_UMASS3 | DEV_UMASS4 | \
-DEV_UMASS5 | DEV_UMASS6 | DEV_UMASS7)
+DEV_UMASS5 | DEV_UMASS6 | DEV_UMASS7 | DEV_UMASS8)
 };
 
 static const struct umsm_type umsm_devs[] = {
@@ -154,6 +155,9 @@ static const struct umsm_type umsm_devs[
 
{{ USB_VENDOR_KYOCERA2, USB_PRODUCT_KYOCERA2_KPC650 }, 0},
 
+   {{ USB_VENDOR_MEDIATEK, USB_PRODUCT_MEDIATEK_UMASS_INSTALLER }, 
DEV_UMASS8},
+   {{ USB_VENDOR_MEDIATEK, USB_PRODUCT_MEDIATEK_DC_4COM }, 0},
+
/* XXX Some qualcomm devices are missing */
{{ USB_VENDOR_QUALCOMM, USB_PRODUCT_QUALCOMM_MSM_DRIVER }, DEV_UMASS1},
{{ USB_VENDOR_QUALCOMM, USB_PRODUCT_QUALCOMM_MSM_DRIVER2 }, DEV_UMASS4},
@@ -303,6 +307,11 @@ umsm_match(struct device *parent, void *
return UMATCH_VENDOR_IFACESUBCLASS;
} else
return UMATCH_NONE;
+   /* Ignore first two (MBIM?) interfaces of 
USB_PRODUCT_MEDIATEK_DC_4COM */
+   } else if (uaa-vendor == USB_VENDOR_MEDIATEK 
+  uaa-product == USB_PRODUCT_MEDIATEK_DC_4COM 
+  (uaa-ifaceno == 0 || uaa-ifaceno == 1)) {
+   return UMATCH_NONE;
} else
return UMATCH_VENDOR_IFACESUBCLASS;
}
@@ -703,6 +712,13 @@ umsm_umass_changemode(struct umsm_softc 
cbw.bCBWFlags = CBWFLAGS_IN;
cbw.CBWCDB[0] = UMASS_SERVICE_ACTION_OUT;
cbw.CBWCDB[1] = 0x03;
+   break;
+   case DEV_UMASS8:
+   USETDW(cbw.dCBWDataTransferLength, 0x0);
+   cbw.bCBWFlags = CBWFLAGS_OUT;
+   cbw.CBWCDB[0] = 0xf0;
+   cbw.CBWCDB[1] = 0x01;
+   cbw.CBWCDB[2] = 0x03;
break;
default:
DPRINTF((%s: unknown device type.\n, 

Re: MediaTek Inc. UMTS stick

[Ingo Feinerer]

On Mon, Jan 05, 2015 at 02:03:38PM +0100, Martin Pieuchot wrote:
 Thanks for your diff.  No need to cross lists, tech@ is the good place
 for a diff :)

OK :)

 On 05/01/15(Mon) 13:08, Ingo Feinerer wrote:
  Please find attached at the end of this mail a diff for the MediaTek
  Inc. DC_4COM (that is the product name that the Linux option driver
  uses:
  https://raw.githubusercontent.com/torvalds/linux/master/drivers/usb/serial/option.c)
  UMTS USB stick (= MEDION S4222). It solves the problems described at
  http://article.gmane.org/gmane.os.openbsd.misc/219029 . This gives me
  
  umsm0 at uhub7 port 1 configuration 1 interface 0 MediaTek Inc Product 
  rev 2.00/3.00 addr 4
  umsm0 detached
  umsm0 at uhub7 port 1 configuration 1 interface 2 MediaTek Inc Product 
  rev 2.00/3.00 addr 4
  ucom0 at umsm0
  umsm1 at uhub7 port 1 configuration 1 interface 3 MediaTek Inc Product 
  rev 2.00/3.00 addr 4
  ucom1 at umsm1
  umsm2 at uhub7 port 1 configuration 1 interface 4 MediaTek Inc Product 
  rev 2.00/3.00 addr 4
  ucom2 at umsm2
  umsm3 at uhub7 port 1 configuration 1 interface 5 MediaTek Inc Product 
  rev 2.00/3.00 addr 4
  ucom3 at umsm3
  umass0 at uhub7 port 1 configuration 1 interface 6 MediaTek Inc Product 
  rev 2.00/3.00 addr 4
  umass0: using SCSI over Bulk-Only
  scsibus4 at umass0: 2 targets, initiator 0
  sd1 at scsibus4 targ 1 lun 0: MEDIATEK, FLASH DISK, 6225 SCSI0 0/direct 
  removable
  ugen0 at uhub7 port 1 configuration 1 MediaTek Inc Product rev 2.00/3.00 
  addr 4
  
  and ucom0 (cuaU0) now works as expected (successfully tested with cu and 
  pppd).
  
  Please note that I had to ignore the first two interfaces when umsm
  attaches (otherwise ucom[0-3] do not work although they attach).
 
 How did you find this out?

Trial and error.

 The linux kernel seems to filter base on the interface subclass and
 protocol.  I'd prefer if you could do the same, as filtering on
 interface index is fragile.  Bad things can happen if the device
 number is reused for example.

True. Please find attached a revised diff. The interface class,
subclass, and protocol values now directly match those in
https://raw.githubusercontent.com/torvalds/linux/master/drivers/usb/serial/option.c
and coincide with my observations.

Successfully tested on amd64.

In general I think it would be more elegant to have a table of
whitelisted interfaces instead of hardcoding the device. However, as
long there are no more special devices the current proposed solution is
more succinct.

 One more comment:
 
  +/* MediaTek Inc. products */
  +product MEDIATEK UMASS_INSTALLER   0x0002  USB installer
  +product MEDIATEK DC_4COM   0x00a5  UMTS USB modem
 
 I'd simply use MEDIATEK UMASS this way you won't exceed 80 columns in
 umsn.c.

Done.

This diff also includes a more detailed comment explaining that this
whitelist stunt is needed (requested by Stuart Henderson).

Thanks a lot for your input!

Best regards,
Ingo

Index: sys/dev/usb/usbdevs
===
RCS file: /cvs/src/sys/dev/usb/usbdevs,v
retrieving revision 1.641
diff -u -p -r1.641 usbdevs
--- sys/dev/usb/usbdevs 4 Dec 2014 10:41:42 -   1.641
+++ sys/dev/usb/usbdevs 5 Jan 2015 18:38:00 -
@@ -452,6 +452,7 @@ vendor CTC  0x0e5e  CONWISE Technology
 vendor HAWKING 0x0e66  Hawking
 vendor FOSSIL  0x0e67  Fossil
 vendor GMATE   0x0e7e  G.Mate
+vendor MEDIATEK0x0e8d  MediaTek Inc.
 vendor OTI 0x0ea0  Ours Technology
 vendor PILOTECH0x0eaf  Pilotech
 vendor NOVATECH0x0eb0  Nova Tech
@@ -2824,6 +2825,10 @@ product MCT ML_4500  0x0302  ML-4500
 
 /* MDS products */
 product MDS ISDBT  0x0001  MDS ISDB-T tuner
+
+/* MediaTek Inc. products */
+product MEDIATEK UMASS 0x0002  USB MSM installer
+product MEDIATEK DC_4COM   0x00a5  UMTS USB modem
 
 /* MEI products */
 product MEI CASHFLOW_SC0x1100  Cashflow-SC
Index: sys/dev/usb/umsm.c
===
RCS file: /cvs/src/sys/dev/usb/umsm.c,v
retrieving revision 1.100
diff -u -p -r1.100 umsm.c
--- sys/dev/usb/umsm.c  12 Jul 2014 21:24:33 -  1.100
+++ sys/dev/usb/umsm.c  5 Jan 2015 18:38:00 -
@@ -109,8 +109,9 @@ struct umsm_type {
 #defineDEV_UMASS5  0x0100
 #defineDEV_UMASS6  0x0200
 #defineDEV_UMASS7  0x0400
+#defineDEV_UMASS8  0x1000
 #define DEV_UMASS  (DEV_UMASS1 | DEV_UMASS2 | DEV_UMASS3 | DEV_UMASS4 | \
-DEV_UMASS5 | DEV_UMASS6 | DEV_UMASS7)
+DEV_UMASS5 | DEV_UMASS6 | DEV_UMASS7 | DEV_UMASS8)
 };
 
 static const struct umsm_type umsm_devs[] = {
@@ -154,6 +155,9 @@ static const struct umsm_type umsm_devs[
 
{{ USB_VENDOR_KYOCERA2, USB_PRODUCT_KYOCERA2_KPC650 }, 0},
 
+   {{ USB_VENDOR_MEDIATEK, USB_PRODUCT_MEDIATEK_UMASS }, DEV_UMASS8},
+   {{ USB_VENDOR_MEDIATEK