Dear Maintainer,

   * What led up to the situation?

   I upgraded my kernel version. To be clear, the issue first presented itself sometime around kernel version 6.15.

   * What exactly did you do (or not do) that was effective (or
     ineffective)?

   When booting my machine, ACPI BIOS ERROR's appear, showing that Field D008 exceeds the target buffer size. I believe this causes something in the ACPI to act as if my machine's wifi kill switch is permanently on.

   * What was the outcome of this action?

   The wlan interface of my machine is unusable and cannot be unblocked using rfkill or setting kernel boot options

   * What outcome did you expect instead?

   The wifi to work.

So, specifically this is the error's I'm seeing on boot:

_AML_BUFFER_LIMIT, Field [D008] at bit offset/length 128/8 exceeds size of targe>
od \_SB.WMID.HWMC due to previous error (AE_AML_BUFFER_LIMIT) 20250807/psparse-5>
od \_SB.WMID.WMAD due to previous error (AE_AML_BUFFER_LIMIT) 20250807/psparse-5>

My own research points to this being an error in the dsdt section of ACPI.

I have my acpidump information - if this is useful then just let me know.

I've also seen a very similar looking issue get resolved in this thread: https://bugzilla.redhat.com/show_bug.cgi?id=1520703

This is the kernel announcement containing the fix for the bug in that thread: https://lwn.net/Articles/806395/

Many thanks,

Josh

Just seen that the error messages I pasted into my initial bug report were clipped. Here are the errors in full:

[   35.029627] ACPI BIOS Error (bug): AE_AML_BUFFER_LIMIT, Field [D008] at bit offset/length 128/8 exceeds size of target Buffer (128 bits) (20250807/dsopcode-198)
[   35.029663] ACPI Error: Aborting method \_SB.WMID.HWMC due to previous error (AE_AML_BUFFER_LIMIT) 20250807/psparse-529)
[   35.029688] ACPI Error: Aborting method \_SB.WMID.WMAD due to previous error (AE_AML_BUFFER_LIMIT) 20250807/psparse-529)

Thanks,

Josh

Hi,

Em qua., 14 de out. de 2026, 22:27, Josh Burke <josh@joshburke.co.uk>
escreveu:
and HP laptops) attempts to write 8 bits starting at an offset of 128
within a 128-bit buffer—that is, one byte beyond its limit. The kernel
interrupts only this specific method and proceeds with the boot process as
usual. Often, these messages can be eliminated through adjustments. A
128-bit buffer is equivalent to a 16-byte block of memory (128 ÷ 8), with
bits numbered from 0 to 127.
Field D008 at offset 128, with a length of 8 bits: the firmware has
declared a field named D008 that begins at bit 128 and occupies 8 bits (1
byte). However, bit 128 is the very first bit beyond the end of the buffer,
which concludes at bit 127. Consequently, the entire field lies outside the
buffer.
An analogy: imagine a drawer with 16 compartments numbered 0 to 15,
accompanied by an instruction to "place the object in compartment 16." That
compartment does not exist.
Why the kernel interrupts execution: the Linux ACPI interpreter checks
these boundaries before accessing memory. Upon detecting that a field
extends beyond the buffer, it returns `AE_AML_BUFFER_LIMIT` and cancels the
method currently being executed (HWMC, as well as the method that called
it, WMAD). Since the remainder of the boot process does not depend on this,
it continues normally. Therefore, this is typically an issue related to the
BIOS.

Leandro