ibmvtpm_send_crq(9)

💅 ibmvtpm_send_crq(9) — like, the Kernel API Manual, section 9 bestie ✨

okay so real talk, this is <hotpink>ibmvtpm_send_crq</hotpink> from the Kernel API Manual (that's the 9 section, iykyk) and it's genuinely the moment you send a CRQ request. CRQ!! iconic acronym. let's get into it 🫶

---

🌸 NAME 🌸

ibmvtpm_send_crq — send a CRQ request 💌

---

✨ SYNOPSIS ✨

int ibmvtpm_send_crq (struct vio_dev *vdev , u8 valid , u8 msg , u16 len
, u32 data );

---

🎀 ARGUMENTS 🎀

• vdev — the vio device struct 👑 she's the main character here
• valid — the Valid field ✨
• msg — the Type field 🏷️
• len — the Length field 📏
• data — the Data field 💾

---

🦋 DESCRIPTION 🦋

bestie. the ibmvtpm crq is defined like this, and I need you to look at it with me, okay:

Byte  |   0   |   1   |   2   |   3   |   4   |   5   |   6   |   7
-----------------------------------------------------------------------
Word0 | Valid | Type  |     Length    |              Data
-----------------------------------------------------------------------
Word1 |                Reserved
-----------------------------------------------------------------------

which is literally a perfect match for the following structure (on a bigendian host):

struct ibmvtpm_crq { u8 valid; u8 msg; __be16 len; __be32 data; __be64
reserved; } __attribute__((packed, aligned(8)));

BUT — and this is the part that's giving plot twist 🌀 — the value gets passed in a register, so you just compute the numeric value to load into that register and skip the byteswap altogether. no byteswapping, we're not doing that today 💅

endian only affects memory loads and stores, fr fr — registers are internally represented the same way. so like, the endianness drama simply does not apply up in here 😌

---

🔥 RETURN 🔥

• 0 (H_SUCCESS) — success, slay, we ate 🎉
• Non-zero — failure, it's giving sad girl hours 😭

---

🌈 SEE ALSO 🌈

kernel file ./drivers/char/tpm/tpm_ibmvtpm.c plus the whole friend group:

ibmvtpm_send_crq_word(9), tpm_ibmvtpm_recv(9), ibmvtpm_crq_send_init(9), tpm_ibmvtpm_resume(9), tpm_ibmvtpm_send(9), ibmvtpm_crq_get_rtce_size(9), ibmvtpm_crq_get_version(9), ibmvtpm_crq_send_init_complete(9), tpm_ibmvtpm_remove(9), tpm_ibmvtpm_get_desired_dma(9), tpm_ibmvtpm_suspend(9), ibmvtpm_reset_crq(9), ibmvtpm_crq_get_next(9), ibmvtpm_crq_process(9), ibmvtpm_interrupt(9), tpm_ibmvtpm_probe(9), ibmvtpm_module_init(9), ibmvtpm_module_exit(9)

---

💀 real one heads up: everything above is the actual spec, I'm just being loud about it. don't @ me if you byteswap when you shouldn't have.

okay that's the whole page, mwah 🫶 so this little love letter lives in ./drivers/char/tpm and it's stamped October 2026 — go forth and send your CRQs, periodt 💖🧚‍♀️🌸