Re: [PATCH v2] MIPS: Fix assembly error from MIPSr2 code used within MIPS_ISA_ARCH_LEVEL

[Thomas Bogendoerfer]

On Fri, Oct 22, 2021 at 12:58:23AM +0200, Maciej W. Rozycki wrote:
> Fix assembly errors like:
> 
> {standard input}: Assembler messages:
> {standard input}:287: Error: opcode not supported on this processor: mips3 (mips3) `dins $10,$7,32,32'
> {standard input}:680: Error: opcode not supported on this processor: mips3 (mips3) `dins $10,$7,32,32'
> {standard input}:1274: Error: opcode not supported on this processor: mips3 (mips3) `dins $12,$9,32,32'
> {standard input}:2175: Error: opcode not supported on this processor: mips3 (mips3) `dins $10,$7,32,32'
> make[1]: *** [scripts/Makefile.build:277: mm/highmem.o] Error 1
> 
> with code produced from `__cmpxchg64' for MIPS64r2 CPU configurations 
> using CONFIG_32BIT and CONFIG_PHYS_ADDR_T_64BIT.
> 
> This is due to MIPS_ISA_ARCH_LEVEL downgrading the assembly architecture 
> to `r4000' i.e. MIPS III for MIPS64r2 configurations, while there is a 
> block of code containing a DINS MIPS64r2 instruction conditionalized on 
> MIPS_ISA_REV >= 2 within the scope of the downgrade.
> 
> The assembly architecture override code pattern has been put there for 
> LL/SC instructions, so that code compiles for configurations that select 
> a processor to build for that does not support these instructions while 
> still providing run-time support for processors that do, dynamically 
> switched by non-constant `cpu_has_llsc'.  It went in with linux-mips.org 
> commit aac8aa7717a2 ("Enable a suitable ISA for the assembler around 
> ll/sc so that code builds even for processors that don't support the 
> instructions. Plus minor formatting fixes.") back in 2005.
> 
> Fix the problem by wrapping these instructions along with the adjacent 
> SYNC instructions only, following the practice established with commit 
> cfd54de3b0e4 ("MIPS: Avoid move psuedo-instruction whilst using 
> MIPS_ISA_LEVEL") and commit 378ed6f0e3c5 ("MIPS: Avoid using .set mips0 
> to restore ISA").  Strictly speaking the SYNC instructions do not have 
> to be wrapped as they are only used as a Loongson3 erratum workaround, 
> so they will be enabled in the assembler by default, but do this so as 
> to keep code consistent with other places.
> 
> Reported-by: kernel test robot <lkp@xxxxxxxxx>
> Signed-off-by: Maciej W. Rozycki <macro@xxxxxxxxxxx>
> Fixes: c7e2d71dda7a ("MIPS: Fix set_pte() for Netlogic XLR using cmpxchg64()")
> Cc: stable@xxxxxxxxxxxxxxx # v5.1+
> ---
> So it's turned out I was right after all with v1, just did not correctly 
> remember the reason for the arrangement with an ISA override around LL/SC, 
> actually made by myself back in 2005.  Therefore here is a minor update to 
> v1 only, which was actually right except for a cosmetic issue around SYNC.
> 
> And as I have thought major decluttering is due here.  What seemed a good 
> idea back in 2005 has never materialised and I think we don't really want 
> to support dynamic `cpu_has_llsc' anymore in 2021.  But let's leave it for 
> another occasion and fix the build error at hand only.
> 
> Changes from v1:
> 
> - Keep __SYNC references under MIPS_ISA_ARCH_LEVEL.
> 
> - Expand change description.
> ---
>  arch/mips/include/asm/cmpxchg.h |    5 ++++-
>  1 file changed, 4 insertions(+), 1 deletion(-)
> 
> linux-mips-cmpxchg64-isa-arch-level.diff
> Index: linux-test/arch/mips/include/asm/cmpxchg.h
> ===================================================================
> --- linux-test.orig/arch/mips/include/asm/cmpxchg.h
> +++ linux-test/arch/mips/include/asm/cmpxchg.h
> @@ -249,6 +249,7 @@ static inline unsigned long __cmpxchg64(
>  	/* Load 64 bits from ptr */
>  	"	" __SYNC(full, loongson3_war) "		\n"
>  	"1:	lld	%L0, %3		# __cmpxchg64	\n"
> +	"	.set	pop				\n"
>  	/*
>  	 * Split the 64 bit value we loaded into the 2 registers that hold the
>  	 * ret variable.
> @@ -276,12 +277,14 @@ static inline unsigned long __cmpxchg64(
>  	"	or	%L1, %L1, $at			\n"
>  	"	.set	at				\n"
>  #  endif
> +	"	.set	push				\n"
> +	"	.set	" MIPS_ISA_ARCH_LEVEL "		\n"
>  	/* Attempt to store new at ptr */
>  	"	scd	%L1, %2				\n"
>  	/* If we failed, loop! */
>  	"\t" __SC_BEQZ "%L1, 1b				\n"
> -	"	.set	pop				\n"
>  	"2:	" __SYNC(full, loongson3_war) "		\n"
> +	"	.set	pop				\n"
>  	: "=&r"(ret),
>  	  "=&r"(tmp),
>  	  "=" GCC_OFF_SMALL_ASM() (*(unsigned long long *)ptr)

applied to mips-next.

Thomas.

-- 
Crap can work. Given enough thrust pigs will fly, but it's not necessarily a
good idea.                                                [ RFC1925, 2.3 ]