| Message ID | 20211220211130.88590-12-pavel.hofman@ivitera.com |
|---|---|
| State | New |
| Headers | show |
| Series | usb: gadget: audio: Multiple rates, dyn. bInterval | expand |
On Mon, Dec 20, 2021 at 10:11:30PM +0100, Pavel Hofman wrote: > So far bInterval for HS and SS was fixed at 4, disallowing faster > samplerates. The patch determines the largest bInterval (4 to 1) for > which the required bandwidth of the max samplerate fits the max allowed > packet size. If the required bandwidth exceeds max bandwidth for > single-packet mode (ep->mc=1), bInterval is left at 1. I'm not sure if this is desirable - there are more concerns around the interval than just whether the bandwidth is available. The nice thing about having the HS/SS interval at 4 when the FS value is 1 is that these both correspond to 1ms, which means the calculations for minimum buffer & period sizes are the same for FS/HS/SS. How do FS transfers work if the bandwidth requirements necessitate a smaller interval for HS/SS? Doesn't that mean the FS transfers must be too big? I don't think there has ever been a check that the configured sample size, channel count and interval actually fit in the max packet size for an endpoint. Is that something that should be checked to give an error on bind if the configuration can't work? > The FS mode is left at fixed bInterval=1. > > Signed-off-by: Pavel Hofman <pavel.hofman@ivitera.com>
Dne 21. 12. 21 v 13:29 John Keeping napsal(a): > On Mon, Dec 20, 2021 at 10:11:30PM +0100, Pavel Hofman wrote: >> So far bInterval for HS and SS was fixed at 4, disallowing faster >> samplerates. The patch determines the largest bInterval (4 to 1) >> for which the required bandwidth of the max samplerate fits the >> max allowed packet size. If the required bandwidth exceeds max >> bandwidth for single-packet mode (ep->mc=1), bInterval is left at >> 1. > > I'm not sure if this is desirable - there are more concerns around > the interval than just whether the bandwidth is available. > > The nice thing about having the HS/SS interval at 4 when the FS > value is 1 is that these both correspond to 1ms, which means the > calculations for minimum buffer & period sizes are the same for > FS/HS/SS. Please do you see any specific place in u_audio.c where the interval of 1ms is assumed? * Buffer/period size max limits are fixed * Bufer min size is calculated from the max_packet_size * snd_pcm_period_elapsed() is called when the current request fill overlaps the period boundary: if ((hw_ptr % snd_pcm_lib_period_bytes(substream)) < req->actual) snd_pcm_period_elapsed(substream); The fixed HS bInterval=4 severely limits the available bandwidth, disallowing even the very basic 192kHz/2ch/24bits config. In f_uac2.c both HS/SS the max packet size, async EP OUT feedback value, as well as async EP IN momentary packet size calculations already take into account the bInterval of the respective endpoint. I have been using bInterval < 4 in most of my tests for almost a year, testing packet sizes at up to 1024 bytes per 125us uframe, both directions, and the gadget has been bitperfect for samplerates up to 4MHz (including correctly working async feedback, tested on linux (up to 4MHz) and windows 10 WASAPI exclusive (up to 1.5MHz). For larger samplerates tests I increased the buffers like in the patch below but I did it just in case to minimize probability of xruns. It's not part of this patchset and should be configured dynamically too, if actually needed at all: diff --git a/drivers/usb/gadget/function/u_audio.c b/drivers/usb/gadget/function/u_audio.c index 58e18952953b..1dedbf324141 100644 --- a/drivers/usb/gadget/function/u_audio.c +++ b/drivers/usb/gadget/function/u_audio.c @@ -23,8 +23,8 @@ #include "u_audio.h" -#define BUFF_SIZE_MAX (PAGE_SIZE * 16) -#define PRD_SIZE_MAX PAGE_SIZE +#define BUFF_SIZE_MAX (PAGE_SIZE * 16 * 64) +#define PRD_SIZE_MAX PAGE_SIZE * 64 #define MIN_PERIODS 4 enum { diff --git a/drivers/usb/gadget/function/u_uac2.h b/drivers/usb/gadget/function/u_uac2.h index 8058217322f8..ae485913db46 100644 --- a/drivers/usb/gadget/function/u_uac2.h +++ b/drivers/usb/gadget/function/u_uac2.h @@ -30,7 +30,7 @@ #define UAC2_DEF_MAX_DB 0 /* 0 dB */ #define UAC2_DEF_RES_DB (1*256) /* 1 dB */ -#define UAC2_DEF_REQ_NUM 2 +#define UAC2_DEF_REQ_NUM 8 #define UAC2_DEF_INT_REQ_NUM 10 struct f_uac2_opts { > > How do FS transfers work if the bandwidth requirements necessitate a > smaller interval for HS/SS? Doesn't that mean the FS transfers > must be too big? Only UAC2 HS/SS bIntervals are dynamic with this patch, FS stays fixed at 1ms. For HS/SS the max packet size is calculated together with the bInterval, so that the largest bInterval possible to fit the ISOC max packetsize limits is chosen. > > I don't think there has ever been a check that the configured sample > size, channel count and interval actually fit in the max packet > size for an endpoint. Is that something that should be checked to > give an error on bind if the configuration can't work? The existing code has never had checks for any of that. Actually the dynamic bInterval calculation in this patch handles the bInterval and packetsize for configured parameters up to maximum ISOC bandwidth. Next version of this patch will at least warn about exceeding the overall available bandwidth. There are many patches to go before the audio gadget becomes fool-proof, but at least it should be practically usable with these patches (when finalized) and the gaudio controller example implementation. I will send the new patchset version with notes for each patch hopefully tomorrow. Thanks a lot for your help, Pavel.
On Wed, 22 Dec 2021 14:35:07 +0100 Pavel Hofman <pavel.hofman@ivitera.com> wrote: > Dne 21. 12. 21 v 13:29 John Keeping napsal(a): > > On Mon, Dec 20, 2021 at 10:11:30PM +0100, Pavel Hofman wrote: > >> So far bInterval for HS and SS was fixed at 4, disallowing faster > >> samplerates. The patch determines the largest bInterval (4 to 1) > >> for which the required bandwidth of the max samplerate fits the > >> max allowed packet size. If the required bandwidth exceeds max > >> bandwidth for single-packet mode (ep->mc=1), bInterval is left at > >> 1. > > > > I'm not sure if this is desirable - there are more concerns around > > the interval than just whether the bandwidth is available. > > > > The nice thing about having the HS/SS interval at 4 when the FS > > value is 1 is that these both correspond to 1ms, which means the > > calculations for minimum buffer & period sizes are the same for > > FS/HS/SS. > > Please do you see any specific place in u_audio.c where the interval of > 1ms is assumed? > > * Buffer/period size max limits are fixed > * Bufer min size is calculated from the max_packet_size > * snd_pcm_period_elapsed() is called when the current request fill > overlaps the period boundary: > > if ((hw_ptr % snd_pcm_lib_period_bytes(substream)) < req->actual) > snd_pcm_period_elapsed(substream); > > > The fixed HS bInterval=4 severely limits the available bandwidth, > disallowing even the very basic 192kHz/2ch/24bits config. Yes, but the problem is if the device enumerates as full-speed the capability is no longer there. I agree that is unlikely to be a problem in real use, but I think it deserves consideration. For the last few years I've been using bInterval == 1 but I also have a hack to disable full-speed operation completely. In my case this is because I want to minimise latency and with the 1ms interval for FS the minimum ALSA period size is too large. Basically, I agree with wanting a smaller bInterval, but I want it for a different reason and I'd like to see a patch that addresses both our use cases ;-) > In f_uac2.c both HS/SS the max packet size, async EP OUT feedback value, > as well as async EP IN momentary packet size calculations already take > into account the bInterval of the respective endpoint. > > I have been using bInterval < 4 in most of my tests for almost a year, > testing packet sizes at up to 1024 bytes per 125us uframe, both > directions, and the gadget has been bitperfect for samplerates up to > 4MHz (including correctly working async feedback, tested on linux (up to > 4MHz) and windows 10 WASAPI exclusive (up to 1.5MHz). For larger > samplerates tests I increased the buffers like in the patch below but I > did it just in case to minimize probability of xruns. It's not part of > this patchset and should be configured dynamically too, if actually > needed at all: This is another case of a different trade-off - I use PREEMPT_RT to minimise xruns and run with a period of 16 samples. > > How do FS transfers work if the bandwidth requirements necessitate a > > smaller interval for HS/SS? Doesn't that mean the FS transfers > > must be too big? > > Only UAC2 HS/SS bIntervals are dynamic with this patch, FS stays fixed > at 1ms. For HS/SS the max packet size is calculated together with the > bInterval, so that the largest bInterval possible to fit the ISOC max > packetsize limits is chosen. I'd really like to see FS mode become unsupported when the packet size is too big. This is a slight issue right now (for 1023 vs 1024) but this patch makes it significantly worse for the high bandwidth case. Right now I have this patch which is a hack but does at least result in an error for the host when trying to enable audio at FS. It would be really nice to properly handle this in the composite gadget core so that the audio function is exposed only at HS/SS with proper DT_OTHER_SPEED_CONFIG handling, but currently that code assumes that the same number of descriptors is provided for each speed. -- 8< -- diff --git a/drivers/usb/gadget/function/f_uac2.c b/drivers/usb/gadget/function/f_uac2.c index 36fa6ef0581b..b4946409b38a 100644 --- a/drivers/usb/gadget/function/f_uac2.c +++ b/drivers/usb/gadget/function/f_uac2.c @@ -1356,6 +1356,9 @@ afunc_set_alt(struct usb_function *fn, unsigned intf, unsigned alt) return 0; } + if (gadget->speed < USB_SPEED_HIGH && alt) + return -EINVAL; + if (intf == uac2->as_out_intf) { uac2->as_out_alt = alt; -- >8 -- > > I don't think there has ever been a check that the configured sample > > size, channel count and interval actually fit in the max packet > > size for an endpoint. Is that something that should be checked to > > give an error on bind if the configuration can't work? > > The existing code has never had checks for any of that. Actually the > dynamic bInterval calculation in this patch handles the bInterval and > packetsize for configured parameters up to maximum ISOC bandwidth. Next > version of this patch will at least warn about exceeding the overall > available bandwidth. > > There are many patches to go before the audio gadget becomes fool-proof, > but at least it should be practically usable with these patches (when > finalized) and the gaudio controller example implementation. Agreed, and I really appreciate the improvements you're making here. The reason I suggested the new checks here is that it makes a lot of sense if the bInterval value is exposed as part of the configfs interface. It means there's one extra value to set for high bandwidth operation, rather than having it "just work", but I think the latency/bandwidth tradeoffs here mean that there's no way for the kernel to select the right value for all scenarios, so really we need to let the user tell us what they want. Regards, John
Dne 22. 12. 21 v 20:50 John Keeping napsal(a): > On Wed, 22 Dec 2021 14:35:07 +0100 > Pavel Hofman <pavel.hofman@ivitera.com> wrote: > >> Dne 21. 12. 21 v 13:29 John Keeping napsal(a): >>> On Mon, Dec 20, 2021 at 10:11:30PM +0100, Pavel Hofman wrote: >>>> So far bInterval for HS and SS was fixed at 4, disallowing faster >>>> samplerates. The patch determines the largest bInterval (4 to 1) >>>> for which the required bandwidth of the max samplerate fits the >>>> max allowed packet size. If the required bandwidth exceeds max >>>> bandwidth for single-packet mode (ep->mc=1), bInterval is left at >>>> 1. >>> >>> I'm not sure if this is desirable - there are more concerns around >>> the interval than just whether the bandwidth is available. >>> >>> The nice thing about having the HS/SS interval at 4 when the FS >>> value is 1 is that these both correspond to 1ms, which means the >>> calculations for minimum buffer & period sizes are the same for >>> FS/HS/SS. >> >> Please do you see any specific place in u_audio.c where the interval of >> 1ms is assumed? >> >> * Buffer/period size max limits are fixed >> * Bufer min size is calculated from the max_packet_size >> * snd_pcm_period_elapsed() is called when the current request fill >> overlaps the period boundary: >> >> if ((hw_ptr % snd_pcm_lib_period_bytes(substream)) < req->actual) >> snd_pcm_period_elapsed(substream); >> >> >> The fixed HS bInterval=4 severely limits the available bandwidth, >> disallowing even the very basic 192kHz/2ch/24bits config. > > Yes, but the problem is if the device enumerates as full-speed the > capability is no longer there. > > I agree that is unlikely to be a problem in real use, but I think it > deserves consideration. Please can you elaborate more on that? If the device enumerates as FS, it's automatically limited to bInterval=1 fullspeed frame. Not much more to do, IIUC. > > For the last few years I've been using bInterval == 1 but I also have a > hack to disable full-speed operation completely. In my case this is > because I want to minimise latency and with the 1ms interval for FS the > minimum ALSA period size is too large. > > Basically, I agree with wanting a smaller bInterval, but I want it for a > different reason and I'd like to see a patch that addresses both our use > cases ;-) > >> In f_uac2.c both HS/SS the max packet size, async EP OUT feedback value, >> as well as async EP IN momentary packet size calculations already take >> into account the bInterval of the respective endpoint. >> >> I have been using bInterval < 4 in most of my tests for almost a year, >> testing packet sizes at up to 1024 bytes per 125us uframe, both >> directions, and the gadget has been bitperfect for samplerates up to >> 4MHz (including correctly working async feedback, tested on linux (up to >> 4MHz) and windows 10 WASAPI exclusive (up to 1.5MHz). For larger >> samplerates tests I increased the buffers like in the patch below but I >> did it just in case to minimize probability of xruns. It's not part of >> this patchset and should be configured dynamically too, if actually >> needed at all: > > This is another case of a different trade-off - I use PREEMPT_RT to > minimise xruns and run with a period of 16 samples. > >>> How do FS transfers work if the bandwidth requirements necessitate a >>> smaller interval for HS/SS? Doesn't that mean the FS transfers >>> must be too big? >> >> Only UAC2 HS/SS bIntervals are dynamic with this patch, FS stays fixed >> at 1ms. For HS/SS the max packet size is calculated together with the >> bInterval, so that the largest bInterval possible to fit the ISOC max >> packetsize limits is chosen. > > I'd really like to see FS mode become unsupported when the packet size > is too big. This is a slight issue right now (for 1023 vs 1024) but > this patch makes it significantly worse for the high bandwidth case. I am afraid I do not understand what the patch makes worse. For FS it always yields bInterval=1 and the corresponding maxPacketSize, a calculation of which has not been changed by the patch. > > Right now I have this patch which is a hack but does at least result in > an error for the host when trying to enable audio at FS. It would be > really nice to properly handle this in the composite gadget core so that > the audio function is exposed only at HS/SS with proper > DT_OTHER_SPEED_CONFIG handling, but currently that code assumes that the > same number of descriptors is provided for each speed. > > -- 8< -- > diff --git a/drivers/usb/gadget/function/f_uac2.c b/drivers/usb/gadget/function/f_uac2.c > index 36fa6ef0581b..b4946409b38a 100644 > --- a/drivers/usb/gadget/function/f_uac2.c > +++ b/drivers/usb/gadget/function/f_uac2.c > @@ -1356,6 +1356,9 @@ afunc_set_alt(struct usb_function *fn, unsigned intf, unsigned alt) > return 0; > } > > + if (gadget->speed < USB_SPEED_HIGH && alt) > + return -EINVAL; > + > if (intf == uac2->as_out_intf) { > uac2->as_out_alt = alt; > > -- >8 -- > >>> I don't think there has ever been a check that the configured sample >>> size, channel count and interval actually fit in the max packet >>> size for an endpoint. Is that something that should be checked to >>> give an error on bind if the configuration can't work? >> >> The existing code has never had checks for any of that. Actually the >> dynamic bInterval calculation in this patch handles the bInterval and >> packetsize for configured parameters up to maximum ISOC bandwidth. Next >> version of this patch will at least warn about exceeding the overall >> available bandwidth. >> >> There are many patches to go before the audio gadget becomes fool-proof, >> but at least it should be practically usable with these patches (when >> finalized) and the gaudio controller example implementation. > > Agreed, and I really appreciate the improvements you're making here. > > The reason I suggested the new checks here is that it makes a lot of > sense if the bInterval value is exposed as part of the configfs > interface. It means there's one extra value to set for high bandwidth > operation, rather than having it "just work", but I think the > latency/bandwidth tradeoffs here mean that there's no way for the kernel > to select the right value for all scenarios, so really we need to let > the user tell us what they want. OK. IMO it could be easily resolved by having the upper bInterval limit for the largest-fitting bInterval check of my patch configurable by new configfs max_bint, defaulting to the existing value of 4. I would leave the default (4), minimizing CPU load, you would set max_bint=1, minimizing latency. Any max_bint value in between would work, while still having available the automated calculation if lower bint value was required for the given parameters. In addition, the final check dev_warn can be chanched to dev_err + returning EINVAL, providing the discussed sanity check. The check would work for FS as well as for HS/SS. This change could be split to three patches: 1. the automated calculation with fixed max_bint=4 - my current patch, dev_warn if max_size_bw > max_size_ep, max_size_bw limited to max_size_ep, no error, only warning. 2. adding the uac2_opts max_bint, using in set_ep_max_packet_size_bint 3. turning the sanity check warning to failing error: changing the dev_warn in the final check to dev_err+ returning error. So the final version could look like this: static int set_ep_max_packet_size_bint(struct device *dev, const struct f_uac2_opts *uac2_opts, struct usb_endpoint_descriptor *ep_desc, enum usb_device_speed speed, bool is_playback) { u16 max_size_bw, max_size_ep; u8 bint; switch (speed) { case USB_SPEED_FULL: max_size_ep = 1023; // fixed bint = 1; max_size_bw = get_max_bw_for_bint(uac2_opts, bint, 1000, is_playback); break; case USB_SPEED_HIGH: case USB_SPEED_SUPER: max_size_ep = 1024; // checking bInterval from max configured bInterval to 1 if the required bandwidth fits for (bint = uac2_opts->max_bint; bint > 0; --bint) { max_size_bw = get_max_bw_for_bint(uac2_opts, bint, 8000, is_playback); if (max_size_bw <= max_size_ep) break; } break; default: return -EINVAL; } if (max_size_bw > max_size_ep) { dev_err(dev, "Req. maxpcktsize %d at bInterval 1= > max ISOC %d, cannot comply!\n", max_size_bw, max_size_ep); return -EINVAL; } ep_desc->wMaxPacketSize = cpu_to_le16(max_size_bw); ep_desc->bInterval = bint; return 0; }
diff --git a/drivers/usb/gadget/function/f_uac2.c b/drivers/usb/gadget/function/f_uac2.c index 984f757de5a4..e72f6f42e1b7 100644 --- a/drivers/usb/gadget/function/f_uac2.c +++ b/drivers/usb/gadget/function/f_uac2.c @@ -333,7 +333,7 @@ static struct usb_endpoint_descriptor fs_epout_desc = { .bEndpointAddress = USB_DIR_OUT, /* .bmAttributes = DYNAMIC */ /* .wMaxPacketSize = DYNAMIC */ - .bInterval = 1, + /* .bInterval = DYNAMIC */ }; static struct usb_endpoint_descriptor hs_epout_desc = { @@ -342,7 +342,7 @@ static struct usb_endpoint_descriptor hs_epout_desc = { /* .bmAttributes = DYNAMIC */ /* .wMaxPacketSize = DYNAMIC */ - .bInterval = 4, + /* .bInterval = DYNAMIC */ }; static struct usb_endpoint_descriptor ss_epout_desc = { @@ -352,7 +352,7 @@ static struct usb_endpoint_descriptor ss_epout_desc = { .bEndpointAddress = USB_DIR_OUT, /* .bmAttributes = DYNAMIC */ /* .wMaxPacketSize = DYNAMIC */ - .bInterval = 4, + /* .bInterval = DYNAMIC */ }; static struct usb_ss_ep_comp_descriptor ss_epout_desc_comp = { @@ -467,7 +467,7 @@ static struct usb_endpoint_descriptor fs_epin_desc = { .bEndpointAddress = USB_DIR_IN, .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ASYNC, /* .wMaxPacketSize = DYNAMIC */ - .bInterval = 1, + /* .bInterval = DYNAMIC */ }; static struct usb_endpoint_descriptor hs_epin_desc = { @@ -476,7 +476,7 @@ static struct usb_endpoint_descriptor hs_epin_desc = { .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ASYNC, /* .wMaxPacketSize = DYNAMIC */ - .bInterval = 4, + /* .bInterval = DYNAMIC */ }; static struct usb_endpoint_descriptor ss_epin_desc = { @@ -486,7 +486,7 @@ static struct usb_endpoint_descriptor ss_epin_desc = { .bEndpointAddress = USB_DIR_IN, .bmAttributes = USB_ENDPOINT_XFER_ISOC | USB_ENDPOINT_SYNC_ASYNC, /* .wMaxPacketSize = DYNAMIC */ - .bInterval = 4, + /* .bInterval = DYNAMIC */ }; static struct usb_ss_ep_comp_descriptor ss_epin_desc_comp = { @@ -659,29 +659,11 @@ static int get_max_srate(const int *srates) return max_srate; } -static int set_ep_max_packet_size(const struct f_uac2_opts *uac2_opts, - struct usb_endpoint_descriptor *ep_desc, - enum usb_device_speed speed, bool is_playback) +static int get_max_bw_for_binterval(const struct f_uac2_opts *uac2_opts, + u8 binterval, unsigned int factor, bool is_playback) { int chmask, srate, ssize; - u16 max_size_bw, max_size_ep; - unsigned int factor; - - switch (speed) { - case USB_SPEED_FULL: - max_size_ep = 1023; - factor = 1000; - break; - - case USB_SPEED_HIGH: - case USB_SPEED_SUPER: - max_size_ep = 1024; - factor = 8000; - break; - - default: - return -EINVAL; - } + u16 max_size_bw; if (is_playback) { chmask = uac2_opts->p_chmask; @@ -699,14 +681,52 @@ static int set_ep_max_packet_size(const struct f_uac2_opts *uac2_opts, srate = srate * (1000 + uac2_opts->fb_max) / 1000; // updated srate is always bigger, therefore DIV_ROUND_UP always yields +1 max_size_bw = num_channels(chmask) * ssize * - (DIV_ROUND_UP(srate, factor / (1 << (ep_desc->bInterval - 1)))); + (DIV_ROUND_UP(srate, factor / (1 << (binterval - 1)))); } else { // adding 1 frame provision for Win10 max_size_bw = num_channels(chmask) * ssize * - (DIV_ROUND_UP(srate, factor / (1 << (ep_desc->bInterval - 1))) + 1); + (DIV_ROUND_UP(srate, factor / (1 << (binterval - 1))) + 1); } + return max_size_bw; +} + +static int set_ep_max_packet_size_bint(const struct f_uac2_opts *uac2_opts, + struct usb_endpoint_descriptor *ep_desc, + enum usb_device_speed speed, bool is_playback) +{ + u16 max_size_bw, max_size_ep; + u8 binterval; + + switch (speed) { + case USB_SPEED_FULL: + max_size_ep = 1023; + // fixed + binterval = 1; + max_size_bw = get_max_bw_for_binterval(uac2_opts, binterval, 1000, + is_playback); + break; + + case USB_SPEED_HIGH: + case USB_SPEED_SUPER: + max_size_ep = 1024; + // checking bInterval from 4 (= 1ms) to 1 if the required bandwidth fits + for (binterval = 4; binterval > 0; --binterval) { + max_size_bw = get_max_bw_for_binterval(uac2_opts, binterval, 8000, + is_playback); + if (max_size_bw <= max_size_ep) { + // found largest bInterval/max_size_bw fitting max_size_ep + break; + } + } + break; + + default: + return -EINVAL; + } + ep_desc->wMaxPacketSize = cpu_to_le16(min_t(u16, max_size_bw, max_size_ep)); + ep_desc->bInterval = binterval; return 0; } @@ -1119,42 +1139,42 @@ afunc_bind(struct usb_configuration *cfg, struct usb_function *fn) } /* Calculate wMaxPacketSize according to audio bandwidth */ - ret = set_ep_max_packet_size(uac2_opts, &fs_epin_desc, USB_SPEED_FULL, + ret = set_ep_max_packet_size_bint(uac2_opts, &fs_epin_desc, USB_SPEED_FULL, true); if (ret < 0) { dev_err(dev, "%s:%d Error!\n", __func__, __LINE__); return ret; } - ret = set_ep_max_packet_size(uac2_opts, &fs_epout_desc, USB_SPEED_FULL, + ret = set_ep_max_packet_size_bint(uac2_opts, &fs_epout_desc, USB_SPEED_FULL, false); if (ret < 0) { dev_err(dev, "%s:%d Error!\n", __func__, __LINE__); return ret; } - ret = set_ep_max_packet_size(uac2_opts, &hs_epin_desc, USB_SPEED_HIGH, + ret = set_ep_max_packet_size_bint(uac2_opts, &hs_epin_desc, USB_SPEED_HIGH, true); if (ret < 0) { dev_err(dev, "%s:%d Error!\n", __func__, __LINE__); return ret; } - ret = set_ep_max_packet_size(uac2_opts, &hs_epout_desc, USB_SPEED_HIGH, + ret = set_ep_max_packet_size_bint(uac2_opts, &hs_epout_desc, USB_SPEED_HIGH, false); if (ret < 0) { dev_err(dev, "%s:%d Error!\n", __func__, __LINE__); return ret; } - ret = set_ep_max_packet_size(uac2_opts, &ss_epin_desc, USB_SPEED_SUPER, + ret = set_ep_max_packet_size_bint(uac2_opts, &ss_epin_desc, USB_SPEED_SUPER, true); if (ret < 0) { dev_err(dev, "%s:%d Error!\n", __func__, __LINE__); return ret; } - ret = set_ep_max_packet_size(uac2_opts, &ss_epout_desc, USB_SPEED_SUPER, + ret = set_ep_max_packet_size_bint(uac2_opts, &ss_epout_desc, USB_SPEED_SUPER, false); if (ret < 0) { dev_err(dev, "%s:%d Error!\n", __func__, __LINE__);
So far bInterval for HS and SS was fixed at 4, disallowing faster samplerates. The patch determines the largest bInterval (4 to 1) for which the required bandwidth of the max samplerate fits the max allowed packet size. If the required bandwidth exceeds max bandwidth for single-packet mode (ep->mc=1), bInterval is left at 1. The FS mode is left at fixed bInterval=1. Signed-off-by: Pavel Hofman <pavel.hofman@ivitera.com> --- drivers/usb/gadget/function/f_uac2.c | 90 +++++++++++++++++----------- 1 file changed, 55 insertions(+), 35 deletions(-)