Message-ID: <4d498c7d1808e5e9090c43822e7c79447e11bbce.camel@linux.ibm.com>
Subject: Re: 6.2-rc7 fails building on Talos II: memory.c:(.text+0x2e14):
 undefined reference to `hash__tlb_flush'
From:   Benjamin Gray <bgray@linux.ibm.com>
To:     "Erhard F." <erhard_f@mailbox.org>, linuxppc-dev@lists.ozlabs.org
Cc:     linux-kernel@vger.kernel.org
Date:   Thu, 16 Feb 2023 16:56:40 +1100
In-Reply-To: <20230216005535.6bff7aa6@yea>
References: <20230216005535.6bff7aa6@yea>
Content-Type: text/plain; charset="UTF-8"
User-Agent: Evolution 3.46.3 (3.46.3-1.fc37) 
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
Precedence: bulk

On Thu, 2023-02-16 at 00:55 +0100, Erhard F. wrote:
> Just noticed a build failure on 6.2-rc7 for my Talos 2 (.config
> attached):
>=20
> =C2=A0# make
> =C2=A0 CALL=C2=A0=C2=A0=C2=A0 scripts/checksyscalls.sh
> =C2=A0 UPD=C2=A0=C2=A0=C2=A0=C2=A0 include/generated/utsversion.h
> =C2=A0 CC=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 init/version-timestamp.o
> =C2=A0 LD=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 .tmp_vmlinux.kallsyms1
> ld: ld: DWARF error: could not find abbrev number 6
> mm/memory.o: in function `unmap_page_range':
> memory.c:(.text+0x2e14): undefined reference to `hash__tlb_flush'
> ld: memory.c:(.text+0x2f8c): undefined reference to `hash__tlb_flush'
> ld: ld: DWARF error: could not find abbrev number 3117
> mm/mmu_gather.o: in function `tlb_remove_table':
> mmu_gather.c:(.text+0x584): undefined reference to `hash__tlb_flush'
> ld: mmu_gather.c:(.text+0x6c4): undefined reference to
> `hash__tlb_flush'
> ld: mm/mmu_gather.o: in function `tlb_flush_mmu':
> mmu_gather.c:(.text+0x80c): undefined reference to `hash__tlb_flush'
> ld: mm/mmu_gather.o:mmu_gather.c:(.text+0xbe0): more undefined
> references to `hash__tlb_flush' follow
> make[1]: *** [scripts/Makefile.vmlinux:35: vmlinux] Fehler 1
> make: *** [Makefile:1264: vmlinux] Error 2
>=20
> As 6.2-rc6 was good on this machine I did a quick bisect which
> revealed this commit:
>=20
> =C2=A0# git bisect bad
> 1665c027afb225882a5a0b014c45e84290b826c2 is the first bad commit
> commit 1665c027afb225882a5a0b014c45e84290b826c2
> Author: Michael Ellerman <mpe@ellerman.id.au>
> Date:=C2=A0=C2=A0 Tue Jan 31 22:14:07 2023 +1100
>=20
> =C2=A0=C2=A0=C2=A0 powerpc/64s: Reconnect tlb_flush() to hash__tlb_flush()
> =C2=A0=C2=A0=C2=A0=20
> =C2=A0=C2=A0=C2=A0 Commit baf1ed24b27d ("powerpc/mm: Remove empty hash__ =
functions")
> =C2=A0=C2=A0=C2=A0 removed some empty hash MMU flushing routines, but got=
 a bit
> overeager
> =C2=A0=C2=A0=C2=A0 and also removed the call to hash__tlb_flush() from tl=
b_flush().
> =C2=A0=C2=A0=C2=A0=20
> =C2=A0=C2=A0=C2=A0 In regular use this doesn't lead to any noticable brea=
kage, which
> is a
> =C2=A0=C2=A0=C2=A0 little concerning. Presumably there are flushes happen=
ing via
> other
> =C2=A0=C2=A0=C2=A0 paths such as arch_leave_lazy_mmu_mode(), and/or a bit=
 of luck.
> =C2=A0=C2=A0=C2=A0=20
> =C2=A0=C2=A0=C2=A0 Fix it by reinstating the call to hash__tlb_flush().
> =C2=A0=C2=A0=C2=A0=20
> =C2=A0=C2=A0=C2=A0 Fixes: baf1ed24b27d ("powerpc/mm: Remove empty hash__ =
functions")
> =C2=A0=C2=A0=C2=A0 Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
> =C2=A0=C2=A0=C2=A0 Link:
> https://lore.kernel.org/r/20230131111407.806770-1-mpe@ellerman.id.au
>=20
> =C2=A0arch/powerpc/include/asm/book3s/64/tlbflush.h | 2 ++
> =C2=A01 file changed, 2 insertions(+)
>=20
>=20
> Regards,
> Erhard

Looks like the `return` on the radix version wasn't added back, so it
falls through to the hash call too.