linux-5.10.y: ia64 build broken since 5.10.221-rc1

[johnny-mnemonic]

https://github.com/linux-ia64/linux-stable-rc/actions/runs/9771252437/job/26974019958#step:8:3524:

[...]
CC [M]  drivers/pps/pps.o
drivers/firmware/efi/memmap.c:16:10: fatal error: asm/efi.h: No such file or directory
   16 | #include <asm/efi.h>
      |          ^~~~~~~~~~~
compilation terminated.
make[3]: *** [scripts/Makefile.build:286: drivers/firmware/efi/memmap.o] Error 1
make[3]: *** Waiting for unfinished jobs....
  CC [M]  drivers/pps/kapi.o
make[2]: *** [scripts/Makefile.build:503: drivers/firmware/efi] Error 2
make[1]: *** [scripts/Makefile.build:503: drivers/firmware] Error 2
make[1]: *** Waiting for unfinished jobs....
[...]

[johnny-mnemonic]

I suspect a relation to the following change:

gregkh/linux@31e0721#diff-fbfb78e650ae911babf6c78ae5b7537bb6a449d19d116f89b34eea93d83f31acR16

...which adds #include <asm/efi.h> to drivers/firmware/efi/memmap.c.

I don't yet know why we don't hit this issue with other kernel branches, as the actual change in mainline is included in mainline since v6.2-rc1 and most likely also in other linux stable branches. Maybe something else "fixes" the issue but didn't make it to linux-5.10.y?

[johnny-mnemonic]

Indeed, looks like 8ff059b which:

Move[s] some EFI related declarations that are only referenced on IA64 to
a new asm/efi.h arch header.

...was included in linux-5.12.y and upwards, but not below. Let's see if that patch set fixes the build for linux-5.10.y locally...

[johnny-mnemonic]

...was included in linux-5.12.y and upwards, but not below. Let's see if that patch set fixes the build for linux-5.10.y locally...

Confirmed, 8ff059b indeed fixes the build.

[johnny-mnemonic]

GHA build fixed for now with 081c0cc.

Waiting for fix in upstream before closing this issue.