dm_test_mst...c_part2_fail(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… dm_test_mst...c_part2_fail(9) β€” bestie edition πŸŽ€

okay so picture this: it's a page straight out of the Kernel API Manual, and the star of the show is literally dm_test_mst...c_part2_fail(9). yes that name is a whole mouthful, no i will not be shortening it, and yes it ate. let's get into it πŸ¦‹


✨ NAME ✨

dm_test_mst_send_payload_alloc_part2_fail β€” the icon, the moment, the legend: it exercises the failure branch πŸ’€ (on purpose!! that's the whole point!!)


🌸 SYNOPSIS 🌸

void dm_test_mst_send_payload_alloc_part2_fail (struct kunit *test );

like... that's it. that's the whole function. she's void, she takes one arg, she does her little job and leaves. periodt πŸ’…


🫢 ARGUMENTS 🫢

  • test β€” the KUnit test context, aka the VIP pass she needs to actually run. hand her one or she's not doing anything, fr fr 🎟️

πŸ‘‘ DESCRIPTION πŸ‘‘

okay so here's the tea β˜• β€” this one builds a minimal MST topology-state fixture so the helper makes it past all those early NULL checks and actually gets to call drm_dp_add_payload_part2. we love a helper that reaches the interesting part 🌈

and here's the iconic bit: the payload's allocation status is left at its default β€” which, crucial detail, is not DRM_DP_MST_PAYLOAD_ALLOCATION_DFP. so because of that, drm_dp_add_payload_part2 returns -EIO πŸ’€ and it does it without any remote DPCD/AUX traffic. no wire, no chatter, no nothing. quiet queen behavior πŸ™ˆ

then that non-zero return value drives the failure branch, which clears the MST_ALLOCATE_NEW_PAYLOAD status bit. so yeah β€” the test is literally built to make things fail, and that's the point. failure branch gets its moment in the spotlight πŸ”₯

ngl, a test whose entire purpose is to fail is peak main character energy. it's giving "i'm not broken, i'm testing" πŸ’–

πŸ¦‹ SEE ALSO πŸ¦‹

the whole extended friend group, in all their glory:


that's the whole thing, bestie, no notes πŸ’œ

and before you go β€” this lil page lives at ./drivers/gpu/drm/amd/display/... and it's dated October 2026, so like… future icon behavior honestly ⭐ mwah, go run your tests πŸ§šβ€β™€οΈπŸ’–

πŸ“¦ where this page comes from