adf_init_admin_pm(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’– adf_init_admin_pm πŸ’–

okay so bestie, this one's from the Kernel API Manual and honestly? it's a little cutie. it's the function that sends the PM init message to FW β€” that's literally the NAME, no notes. πŸŽ€


✨ SYNOPSIS ✨

int adf_init_admin_pm (struct adf_accel_dev *accel_dev , u32 idle_delay
);

like, she takes two things and gives you back an int. that's the whole vibe check. πŸ’…


🌸 ARGUMENTS 🌸

  • accel_dev β€” pointer to the acceleration device. your main character. the device in question. 🫢
  • idle_delay β€” QAT HW idle time before power gating is initiated, and yes it's a whole little menu of bit patterns, so let's read the menu together, no cap:
    • 000 β€” 64us
    • 001 β€” 128us
    • 010 β€” 256us
    • 011 β€” 512us
    • 100 β€” 1ms
    • 101 β€” 2ms
    • 110 β€” 4ms
    • 111 β€” 8ms

it's giving "three bits, eight moods" ✨ pick your fave wisely.


πŸ¦‹ DESCRIPTION πŸ¦‹

so what she does: the function sends to the FW the admin init message for the PM state configuration.

that's it. that's the whole job. she's the messenger, the FW is the recipient, and the message is the PM state config. iconic behaviour, honestly. πŸ’œ


🌈 RETURN 🌈

  • 0 on success πŸŽ‰
  • error code otherwise 😭 (if you got a nonzero back, something did not slay)

⭐ SEE ALSO ⭐


🧚 little reminder bestie: the idle_delay values above are the actual literal bit patterns from the page β€” don't go improvising your own, iykyk πŸ’€

okay that's the whole page, mwah πŸ’– β€” this was adf_init_admin_pm(9) from the Kernel API Manual, source ./drivers/crypto/intel/qat/qat_common/adf_admin.c (the footer shows ./drivers/crypto/intel/qat/qat...), dated October 2026. go forth and configure your PM state, you gorgeous nerd πŸŒΈπŸ’…

πŸ“¦ where this page comes from