From: Mauro Carvalho Chehab <m.chehab@samsung.com>
To: "Frank Schäfer" <fschaefer.oss@googlemail.com>
Cc: Linux Media Mailing List <linux-media@vger.kernel.org>,
Mauro Carvalho Chehab <mchehab@infradead.org>
Subject: Re: [PATCH 1/4] em28xx: use bInterval on em28xx-audio
Date: Sat, 11 Jan 2014 18:53:13 -0200 [thread overview]
Message-ID: <20140111185313.1355156b@samsung.com> (raw)
In-Reply-To: <52D1491A.1070104@googlemail.com>
Em Sat, 11 Jan 2014 14:37:30 +0100
Frank Schäfer <fschaefer.oss@googlemail.com> escreveu:
> Am 10.01.2014 19:45, schrieb Mauro Carvalho Chehab:
> > Just filling urb->interval with 1 is wrong, and causes a different
> > behaviour with xHCI.
> >
> > With EHCI, the URB size is typically 192 bytes. However, as
> > xHCI specifies intervals in microframes, the URB size becomes
> > too short (24 bytes).
> >
> > With this patch, the interval will be properly initialized, and
> > the device will behave the same if connected into a xHCI or an
> > EHCI device port.
> >
> > Signed-off-by: Mauro Carvalho Chehab <m.chehab@samsung.com>
> > ---
> > drivers/media/usb/em28xx/em28xx-audio.c | 39 ++++++++++++++++++++++++++++++++-
> > 1 file changed, 38 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/media/usb/em28xx/em28xx-audio.c b/drivers/media/usb/em28xx/em28xx-audio.c
> > index 30ee389a07f0..8e6f04873422 100644
> > --- a/drivers/media/usb/em28xx/em28xx-audio.c
> > +++ b/drivers/media/usb/em28xx/em28xx-audio.c
> > @@ -620,10 +620,13 @@ static int em28xx_audio_init(struct em28xx *dev)
> > struct em28xx_audio *adev = &dev->adev;
> > struct snd_pcm *pcm;
> > struct snd_card *card;
> > + struct usb_interface *intf;
> > + struct usb_endpoint_descriptor *e, *ep = NULL;
> > static int devnr;
> > int err, i;
> > const int sb_size = EM28XX_NUM_AUDIO_PACKETS *
> > EM28XX_AUDIO_MAX_PACKET_SIZE;
> > + u8 alt;
> >
> > if (!dev->has_alsa_audio || dev->audio_ifnum < 0) {
> > /* This device does not support the extension (in this case
> > @@ -679,6 +682,34 @@ static int em28xx_audio_init(struct em28xx *dev)
> > em28xx_cvol_new(card, dev, "Surround", AC97_SURROUND_MASTER);
> > }
> >
> > + if (dev->audio_ifnum)
> > + alt = 1;
> > + else
> > + alt = 7;
> > +
> > + intf = usb_ifnum_to_if(dev->udev, dev->audio_ifnum);
> > +
> > + if (intf->num_altsetting <= alt) {
> > + em28xx_errdev("alt %d doesn't exist on interface %d\n",
> > + dev->audio_ifnum, alt);
> > + return -ENODEV;
> > + }
> Hmm... yeah, this the alt setting code looks suspicious.
>
> Take a look at snd_em28xx_capture_open():
>
> if ((dev->alt == 0 || dev->audio_ifnum) && dev->adev.users == 0) {
> if (dev->audio_ifnum)
> dev->alt = 1;
> else
> dev->alt = 7;
> ...
> }
>
> I've been thinking about this for a while and I think this code is based
> on the following assumptions:
> 1.) Video endpoints are always at interface 0
> 2.) Hence, if the audio endpoints are on a separate interface, the
> interface number is > 0
> 3.) Video interfaces always have alt settings 0-7 and 7 is the one with
> the highest bandwith.
> 1.) is definitely wrong, the VAD Laplace webcam for example has video on
> interface #3.
> This needs to be fixed in the core, too.
> Because of that, the (dev->audio_ifnum > 0) check also needs to be fixed
> and dev->is_audio_only should be checked instead.
> 3.) matches what I've seen so far, but seems to be safer to do the same
> what we are doing in em28xx_usb_probe() for the dvb video endpoints.
>
> Whether alt=1 and alt=max are a good choice is a separate question.
alt = 1 for audio is because the audio-only interface on em2861 has only 5
alternates, and alt = 1 has the biggest wMaxPacketSize.
This code can of course be optimized.
alt = <max> for mixed video/audio interface was chosen to avoid the
need of recalculate the video + audio bandwidth.
It would be possible to rework on this code to create a single function
that would adjust the bandwidth considering both video and audio
constraints, but it takes time, and testing. So, when the audio driver
was written, the developer just hardcoded alt = 7, and didn't bother
with that.
Nobody had time to review it so far.
> How many altternate settings does the audio only interface of you em2860
> have and how do they look ?
Interface Descriptor:
bLength 9
bDescriptorType 4
bInterfaceNumber 1
bAlternateSetting 0
bNumEndpoints 1
bInterfaceClass 255 Vendor Specific Class
bInterfaceSubClass 0
bInterfaceProtocol 255
iInterface 0
Endpoint Descriptor:
bLength 7
bDescriptorType 5
bEndpointAddress 0x83 EP 3 IN
bmAttributes 1
Transfer Type Isochronous
Synch Type None
Usage Type Data
wMaxPacketSize 0x0000 1x 0 bytes
bInterval 4
Interface Descriptor:
bLength 9
bDescriptorType 4
bInterfaceNumber 1
bAlternateSetting 1
bNumEndpoints 1
bInterfaceClass 255 Vendor Specific Class
bInterfaceSubClass 0
bInterfaceProtocol 255
iInterface 0
Endpoint Descriptor:
bLength 7
bDescriptorType 5
bEndpointAddress 0x83 EP 3 IN
bmAttributes 1
Transfer Type Isochronous
Synch Type None
Usage Type Data
wMaxPacketSize 0x00c4 1x 196 bytes
bInterval 4
Interface Descriptor:
bLength 9
bDescriptorType 4
bInterfaceNumber 1
bAlternateSetting 2
bNumEndpoints 1
bInterfaceClass 255 Vendor Specific Class
bInterfaceSubClass 0
bInterfaceProtocol 255
iInterface 0
Endpoint Descriptor:
bLength 7
bDescriptorType 5
bEndpointAddress 0x83 EP 3 IN
bmAttributes 1
Transfer Type Isochronous
Synch Type None
Usage Type Data
wMaxPacketSize 0x00b4 1x 180 bytes
bInterval 4
Interface Descriptor:
bLength 9
bDescriptorType 4
bInterfaceNumber 1
bAlternateSetting 3
bNumEndpoints 1
bInterfaceClass 255 Vendor Specific Class
bInterfaceSubClass 0
bInterfaceProtocol 255
iInterface 0
Endpoint Descriptor:
bLength 7
bDescriptorType 5
bEndpointAddress 0x83 EP 3 IN
bmAttributes 1
Transfer Type Isochronous
Synch Type None
Usage Type Data
wMaxPacketSize 0x0084 1x 132 bytes
bInterval 4
Interface Descriptor:
bLength 9
bDescriptorType 4
bInterfaceNumber 1
bAlternateSetting 4
bNumEndpoints 1
bInterfaceClass 255 Vendor Specific Class
bInterfaceSubClass 0
bInterfaceProtocol 255
iInterface 0
Endpoint Descriptor:
bLength 7
bDescriptorType 5
bEndpointAddress 0x83 EP 3 IN
bmAttributes 1
Transfer Type Isochronous
Synch Type None
Usage Type Data
wMaxPacketSize 0x0044 1x 68 bytes
bInterval 4
Interface Descriptor:
bLength 9
bDescriptorType 4
bInterfaceNumber 1
bAlternateSetting 5
bNumEndpoints 1
bInterfaceClass 255 Vendor Specific Class
bInterfaceSubClass 0
bInterfaceProtocol 255
iInterface 0
Endpoint Descriptor:
bLength 7
bDescriptorType 5
bEndpointAddress 0x83 EP 3 IN
bmAttributes 1
Transfer Type Isochronous
Synch Type None
Usage Type Data
wMaxPacketSize 0x0024 1x 36 bytes
bInterval 4
>
> In case of vendor audio endpoints on the same interface as the video
> endpoints, wMaxPacketSize and bInterval seem to be the same for all alt
> settings.
> (The only device I know so far is the HVR-930c: wMaxPacketSize =1x 196
> bytes, bInterval = 4).
Yes, but the interface to be selected, in this case, should estimate both
audio and video, if both are active, in order to select the right
alternate.
In any case, this patch doesn't change anything related to it, and it is
a first step to add a proper code here to detect and select the proper
setup.
> > +
> > + for (i = 0; i < intf->altsetting[alt].desc.bNumEndpoints; i++) {
> > + e = &intf->altsetting[alt].endpoint[i].desc;
> > + if (!usb_endpoint_dir_in(e))
> > + continue;
> > + if (e->bEndpointAddress == EM28XX_EP_AUDIO) {
> > + ep = e;
> > + break;
> > + }
> > + }
> > +
> > + if (!ep) {
> > + em28xx_errdev("Couldn't find an audio endpoint");
> > + return -ENODEV;
> > + }
> > +
> > /* Alloc URB and transfer buffers */
> > for (i = 0; i < EM28XX_AUDIO_BUFS; i++) {
> > struct urb *urb;
> > @@ -707,11 +738,17 @@ static int em28xx_audio_init(struct em28xx *dev)
> > urb->pipe = usb_rcvisocpipe(dev->udev, EM28XX_EP_AUDIO);
> > urb->transfer_flags = URB_ISO_ASAP | URB_NO_TRANSFER_DMA_MAP;
> > urb->transfer_buffer = dev->adev.transfer_buffer[i];
> > - urb->interval = 1;
> > + urb->interval = 1 << (ep->bInterval - 1);
> > urb->complete = em28xx_audio_isocirq;
> > urb->number_of_packets = EM28XX_NUM_AUDIO_PACKETS;
> > urb->transfer_buffer_length = sb_size;
> >
> > + if (!i)
> > + dprintk("Will use ep 0x%02x on intf %d alt %d interval = %d (rcv isoc pipe: 0x%08x)\n",
> > + EM28XX_EP_AUDIO, dev->audio_ifnum, alt,
> > + urb->interval,
> > + urb->pipe);
> > +
> > for (j = k = 0; j < EM28XX_NUM_AUDIO_PACKETS;
> > j++, k += EM28XX_AUDIO_MAX_PACKET_SIZE) {
> > urb->iso_frame_desc[j].offset = k;
>
--
Cheers,
Mauro
next prev parent reply other threads:[~2014-01-11 20:53 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-01-10 18:45 [PATCH 0/4] Fix em28xx audio when using a xHCI USB port Mauro Carvalho Chehab
2014-01-10 18:45 ` [PATCH 1/4] em28xx: use bInterval on em28xx-audio Mauro Carvalho Chehab
2014-01-11 13:37 ` Frank Schäfer
2014-01-11 20:53 ` Mauro Carvalho Chehab [this message]
2014-01-12 15:13 ` Frank Schäfer
2014-01-10 18:45 ` [PATCH 2/4] em28xx-audio: Fix error path Mauro Carvalho Chehab
2014-01-10 18:45 ` [PATCH 3/4] em28xx: don't hardcode audio URB calculus Mauro Carvalho Chehab
2014-01-10 18:45 ` [PATCH 4/4] em28x-audio: fix the period size in bytes Mauro Carvalho Chehab
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20140111185313.1355156b@samsung.com \
--to=m.chehab@samsung.com \
--cc=fschaefer.oss@googlemail.com \
--cc=linux-media@vger.kernel.org \
--cc=mchehab@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.