From: "Cássio Gabriel" <cassiogabrielcontato@gmail.com>
To: Clemens Ladisch <clemens@ladisch.de>,
Jaroslav Kysela <perex@perex.cz>, Takashi Iwai <tiwai@suse.com>
Cc: linux-sound@vger.kernel.org, linux-kernel@vger.kernel.org,
"Cássio Gabriel" <cassiogabrielcontato@gmail.com>
Subject: [PATCH] ALSA: usb-audio: rotate standard MIDI output port scan
Date: Mon, 23 Mar 2026 10:46:24 -0300 [thread overview]
Message-ID: <20260323-usbmidi-port-fairness-v1-1-2d68e97592a1@gmail.com> (raw)
snd_usbmidi_standard_output() iterates output ports in ascending order
and drains each active port until the URB is full. On interfaces where
multiple USB-MIDI cables share one endpoint, sustained traffic on a
lower-numbered port can consume every refill before higher-numbered
ports are even examined.
That behavior dates back to the original implementation and still
applies with the current multi-URB output path. snd_usbmidi_do_output()
can refill several idle URBs in one pass, but each refill restarts the
scan at port 0, so a busy lower-numbered port can keep higher-numbered
ports from making progress at all.
Use ep->current_port as the starting point of the scan and advance it
after each URB fill. This keeps the existing packet formatting and
per-port state handling intact while preventing persistent starvation of
higher-numbered ports.
Signed-off-by: Cássio Gabriel <cassiogabrielcontato@gmail.com>
---
sound/usb/midi.c | 12 ++++++++----
1 file changed, 8 insertions(+), 4 deletions(-)
diff --git a/sound/usb/midi.c b/sound/usb/midi.c
index a8bddc90c0ed..0a5b8941ebda 100644
--- a/sound/usb/midi.c
+++ b/sound/usb/midi.c
@@ -699,15 +699,18 @@ static void snd_usbmidi_transmit_byte(struct usbmidi_out_port *port,
static void snd_usbmidi_standard_output(struct snd_usb_midi_out_endpoint *ep,
struct urb *urb)
{
- int p;
+ int port0 = ep->current_port;
+ int i;
+
+ for (i = 0; i < 0x10; ++i) {
+ int portnum = (port0 + i) & 15;
+ struct usbmidi_out_port *port = &ep->ports[portnum];
- /* FIXME: lower-numbered ports can starve higher-numbered ports */
- for (p = 0; p < 0x10; ++p) {
- struct usbmidi_out_port *port = &ep->ports[p];
if (!port->active)
continue;
while (urb->transfer_buffer_length + 3 < ep->max_transfer) {
uint8_t b;
+
if (snd_rawmidi_transmit(port->substream, &b, 1) != 1) {
port->active = 0;
break;
@@ -715,6 +718,7 @@ static void snd_usbmidi_standard_output(struct snd_usb_midi_out_endpoint *ep,
snd_usbmidi_transmit_byte(port, b, urb);
}
}
+ ep->current_port = (port0 + 1) & 15;
}
static const struct usb_protocol_ops snd_usbmidi_standard_ops = {
---
base-commit: b3c48fa1fb397b490101785ddd87caf2e5513a66
change-id: 20260323-usbmidi-port-fairness-4301533ef31a
Best regards,
--
Cássio Gabriel <cassiogabrielcontato@gmail.com>
next reply other threads:[~2026-03-23 13:46 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-23 13:46 Cássio Gabriel [this message]
2026-03-27 11:45 ` Takashi Iwai
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=20260323-usbmidi-port-fairness-v1-1-2d68e97592a1@gmail.com \
--to=cassiogabrielcontato@gmail.com \
--cc=clemens@ladisch.de \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-sound@vger.kernel.org \
--cc=perex@perex.cz \
--cc=tiwai@suse.com \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®