efx_enqueue_skb_tso(9)

🎀 NAME 🎀

okay so bestie, this is efx_enqueue_skb_tso(9) from the Kernel API Manual — the page whose whole job is to segment and transmit a TSO socket buffer. It's giving "segment + transmit" main character energy ✨ and honestly the whole function is just doing TSO things with a socket buffer. No cap.

✨ SYNOPSIS ✨

int efx_enqueue_skb_tso (struct efx_tx_queue *tx_queue , struct sk_buff
*skb , bool *data_mapped );

🫶 ARGUMENTS 🫶

• tx_queue — the Efx TX queue. That's the queue getting the goods 💅
• skb — the Socket buffer. The buffer in question, periodt.
• data_mapped — Did we map the data? On success, this function always sets it to true ✅

💅 CONTEXT 💅

bestie, vibe check: you must hold netif_tx_lock to call this function. Literally do not skip this lock, fr fr.

🌸 DESCRIPTION 🌸

Bestie, the move is: add socket buffer skb to tx_queue, doing TSO — or return != 0 if skb was not enqueued. That's the whole thing. Lowkey simple, highkey important.

And the ownership tea ☕: skb is consumed unless the return value is EINVAL. So EINVAL means not consumed; every other return value means it got consumed. Not me being dramatic, that's just the rule 😭

🦋 SEE ALSO 🦋

• Kernel file ./drivers/net/ethernet/sfc/tx_tso.c
• struct tso_state(9)
• efx_tx_queue_insert(9)
• tso_fill_packet_with_fragment(9)
• tso_start_new_packet(9)

mwah 💋 this cute lil page's source/version stamp is ./drivers/net/ethernet/sfc, and it's dated October 2026 — efx_enqueue_skb_tso(9) signing off 🫶✨