ps5_control --fan is a silent no-op on PS5 Slim; please expose target_temp #7
Labels
No labels
bug
documentation
duplicate
enhancement
good first issue
help wanted
invalid
question
wontfix
No milestone
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set
Reference
gh-ps5-linux/ps5-linux-tools#7
Loading…
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Summary
On a PS5 Slim (CFI-2015 B01X),
ps5_control --fanreports success while havingno effect whatsoever. Separately, the one fan interface that does work on this
console — the
target_tempsysfs attribute — is not exposed byps5_controlat all.
Environment
linux-ps57.1.3-1ps5-linux-tools,ps5_controlfrom/usr/local/sbink10temp/ Tctl (the only temperature sensor; no SMU on this APU)Bug —
--fanis a silent no-opExit code 0, success message printed, nothing happens. No audible change, no
temperature change, and no kernel log entry when the ioctl fires.
Controlled A/B with system load held constant at idle, 150 samples:
on)max)Settled delta 0.07 °C, which is below the sensor's 0.25 °C quantization
step. The fan is not responding at all.
Likely cause
This is probably downstream of a kernel-side bug rather than a fault in
ps5_controlitself.ICC_FAN_CHANGE_SERVO_PATTERNselects a pattern from theautoservo table, and on this console that table is never populated — the boot
log shows:
I've filed the kernel-side detail separately against
ps5-linux-patches.Regardless of the root cause, the tool should not report success when the
ioctl has no effect. Silent success is what led me to spend two days
concluding the fan was uncontrollable on this hardware. Even just documenting
that the pattern ioctl is a no-op when the servo table is unpopulated would
help.
Minor, related: the usage text advertises only
<on|off>, but the source alsoaccepts
max(pattern 3). Undocumented.Feature request — expose
target_tempThere is a working fan control on this console, and
ps5_controldoesn'texpose it:
This works immediately and reliably — the fan audibly ramps and temperature
tracks the setpoint. Measured on this console with 007 First Light, same
chapter section and settings, boost ON, 2 Hz sampling:
11.7 °C on the mean, 14.2 °C on the peak, with frametimes statistically
unchanged (median 17.6 → 17.9 ms) and no clock-speed cost. The default 91 °C
setpoint leaves only 0.75 °C of headroom under the Zen 2 TjMax of 95 °C.
Suggested:
Since
--fanis what users naturally reach for and it silently does nothing onthis hardware, having the working interface available from the same tool would
make a real difference. Right now nothing in the tooling or documentation points
to
target_temp— I only found it by readinglinux.patch.Kernel-side counterpart filed as ps5-linux/ps5-linux-patches#32 — that one covers why the autoservo table is never populated on this console, which is the likely root cause of the
--fanno-op described above.