dwc2_hsotg_handle_rx(9)

📦 version 7.3.1, in nixpkgs unstable

💅 dwc2_hsotg_handle_rx — kernel API manual, but make it make sense 🎀

okay so bestie, welcome to the glow-up edition of the dwc2_hsotg_handle_rx(9) page from the Kernel API Manual ✨ no cap, this one's short but she's important.

✨ NAME ✨

dwc2_hsotg_handle_rx — the RX FIFO has data 💖 like, actual data. sitting there. waiting for you.

🦋 SYNOPSIS 🦋

void dwc2_hsotg_handle_rx (struct dwc2_hsotg *hsotg );

🌸 ARGUMENTS 🌸

  • hsotg — the device instance 💫 (the main character of this whole operation, honestly)

🧚 DESCRIPTION 🧚

so picture this: the IRQ handler has clocked that the RX FIFO has some data in it that needs processing 👀 so this function's whole job is to go find out what's in there and do the appropriate read. that's it. that's the vibe check.

the RXFIFO is a true FIFO — like, genuinely first-in-first-out, she's not faking it 💅 — and the packets coming out are still in packet chunks. so if you've got x packets received on an endpoint, you'll get x FIFO events delivered, each one with a packet's worth of data in it. one packet, one event. periodt. 🎀

⚠️ okay this part matters, so pay attention bestie: when using DMA, we should not be processing events from the RXFIFO. the actual data should be sent to the memory directly, and instead we turn on the completion interrupts to get notifications of transfer completion. do not skip this, i'm literally begging 🙈

🔥 SEE ALSO 🔥

Kernel file ./drivers/usb/dwc2/gadget.c — that's where she lives 🏠 — plus this absolutely iconic lineup of related functions, all of them (9)-worthy:

and that's the whole page, bestie 🫶 she's short, she's specific, and she will not let you process RXFIFO events while DMA is doing its thing. iykyk.

✨ mwah 💋 this edition of dwc2_hsotg_handle_rx(9) came to you from ./drivers/usb/dwc2, dated October 2026 — stay hydrated, read your FIFOs, and go slay some interrupts 🔥👑

📦 where this page comes from