From 72e09788e46a8ffaaae8009a316915838927af90 Mon Sep 17 00:00:00 2001 From: Boris Dolgov Date: Fri, 15 May 2026 21:28:21 +0200 Subject: rp2040: fix host SET_REPORT (and any OUT-data control xfer) sending DATA0 instead of DATA1 hcd_edpt_xfer() previously reset ep->next_pid to 1 only when the control endpoint direction changed between stages. That handled IN-data control transfers (e.g. GET_REPORT, GET_DESCRIPTOR) where SETUP is OUT and DATA is IN, but not OUT-data class requests like SET_REPORT, where SETUP and DATA are both OUT and the direction-change check is false. ep->next_pid was left at 0 from hcd_edpt_open(), so the DATA stage went on the wire as DATA0 when the device expected DATA1. Strict devices (observed: Elgato Stream Deck) treat this as a protocol violation and disconnect. Key off "endpoint 0" instead of "direction changed", restoring the previous behavior. Interrupt/bulk endpoints take the ep->interrupt_num > 0 branch above and never reach this code, so they are unaffected. --- src/portable/raspberrypi/rp2040/hcd_rp2040.c | 12 +++++++++--- 1 file changed, 9 insertions(+), 3 deletions(-) (limited to 'src') diff --git a/src/portable/raspberrypi/rp2040/hcd_rp2040.c b/src/portable/raspberrypi/rp2040/hcd_rp2040.c index 02a4e055e..064834efb 100644 --- a/src/portable/raspberrypi/rp2040/hcd_rp2040.c +++ b/src/portable/raspberrypi/rp2040/hcd_rp2040.c @@ -617,10 +617,16 @@ bool hcd_edpt_xfer(uint8_t rhport, uint8_t dev_addr, uint8_t ep_addr, uint8_t *b io_rw_32 *buf_reg = dpram_int_ep_buffer_ctrl(ep->interrupt_num); rp2usb_xfer_start(ep, ep_reg, buf_reg, buffer, NULL, buflen); } else { - // Control endpoint can change direction 0x00 <-> 0x80 when changing stages - if (ep_addr != ep->ep_addr) { + // Control transfer data and status stages always start with DATA1, regardless of + // whether the direction changed since the previous stage. SET_REPORT (and any other + // host-to-device class request with an OUT data stage) keeps the same direction + // across SETUP -> DATA, so we cannot key off "direction changed" -- we must reset + // next_pid every time hcd_edpt_xfer is invoked on ep 0. Without this, the data stage + // of SET_REPORT goes out as DATA0 because ep->next_pid is still 0 from hcd_edpt_open(), + // which strict devices treat as a protocol violation and disconnect. + if (tu_edpt_number(ep_addr) == 0) { ep->ep_addr = ep_addr; - ep->next_pid = 1; // data and status stage start with DATA1 + ep->next_pid = 1; } // If EPX is busy with another transfer, mark as pending -- cgit v1.3.1