Hello,
I checked sizeof(sector_t) and it is 8. The assembly code I get is similar
to what Sergey posted.
If it really is a compiler problem, which are the options? Would removing
the inline help?
---
Luiz Angelo Daros de Luca, Me.
[email protected]
2013/4/27 Sergey Vlasov <[email protected]>
> On Sat, Apr 27, 2013 at 02:09:22AM -0300, Luiz Angelo Daros de Luca wrote:
> > The usage of sign-extension might be cause. The problem is that all these
> > variables are sector_t, which as far as I know, is a u64.
> > Even if it was using signed variable, all of them are 64bit and using
> > values much lower that 2^63.
> >
> > As mips is not a 64-bit processor, the compiler must do it by parts. So,
> > where is this problem from? Compiler? Kernel?
>
> Seems that at least gcc-4.6-linaro from AA (mips-openwrt-linux-gcc
> (Linaro GCC 4.6-2012.02) 4.6.3 20120201 (prerelease),
> PKG_MD5SUM:=2b7887846f8e5ac1ca58fe4dfaabf5a6) is buggy. Here is a
> self-contained test case:
>
> madd.c:
>
> =============================================================================
> typedef unsigned long long u64;
> typedef unsigned long u32;
> typedef unsigned char u8;
>
> static inline u32 get_unaligned_le32(const u8 *p)
> {
> return p[0] | p[1] << 8 | p[2] << 16 | p[3] << 24;
> }
>
> static inline u64 my_get_le32(const void *p)
> {
> return (u64)get_unaligned_le32(p);
> }
>
> u64 test_fn(u64 arg1, void *arg2, void *arg3)
> {
> return arg1 + my_get_le32(arg2)*my_get_le32(arg3);
> }
>
> =============================================================================
>
> When I compile this file with:
>
> $STAGING_DIR/bin/mips-openwrt-linux-gcc -march=mips32r2 -S -O2 -Wall -W
> madd.c
>
> I get the following output in madd.s:
>
>
> =============================================================================
> .file 1 "madd.c"
> .section .mdebug.abi32
> .previous
> .gnu_attribute 4, 3
> .abicalls
> .option pic0
> .text
> .align 2
> .globl test_fn
> .set nomips16
> .ent test_fn
> .type test_fn, @function
> test_fn:
> .frame $sp,0,$31 # vars= 0, regs= 0/0, args= 0, gp=
> 0
> .mask 0x00000000,0
> .fmask 0x00000000,0
> .set noreorder
> .set nomacro
> lbu $11,1($6)
> lbu $2,2($6)
> lbu $3,2($7)
> lbu $12,1($7)
> lbu $10,0($6)
> lbu $9,3($6)
> lbu $8,0($7)
> lbu $6,3($7)
> sll $11,$11,8
> sll $7,$12,8
> sll $2,$2,16
> sll $3,$3,16
> or $2,$11,$2
> or $3,$7,$3
> or $2,$2,$10
> sll $7,$9,24
> or $3,$3,$8
> sll $6,$6,24
> mtlo $5
> or $2,$2,$7
> or $3,$3,$6
> mthi $4
> madd $3,$2
> mflo $3
> j $31
> mfhi $2
>
> .set macro
> .set reorder
> .end test_fn
> .size test_fn, .-test_fn
> .ident "GCC: (Linaro GCC 4.6-2012.02) 4.6.3 20120201 (prerelease)"
>
> =============================================================================
>
> The "madd $3,$2" command here is incorrect - it performs sign
> extension of its arguments; it should be "maddu $3,$2".
>
> _______________________________________________
> openwrt-devel mailing list
> [email protected]
> https://lists.openwrt.org/mailman/listinfo/openwrt-devel
>
>
_______________________________________________
openwrt-devel mailing list
[email protected]
https://lists.openwrt.org/mailman/listinfo/openwrt-devel
