qemu-arm
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [PATCH 1/6] hw/arm/aspeed: Fix smpboot[] on big-endian hosts


From: Peter Maydell
Subject: Re: [PATCH 1/6] hw/arm/aspeed: Fix smpboot[] on big-endian hosts
Date: Tue, 3 Jan 2023 17:33:56 +0000

On Thu, 22 Dec 2022 at 21:56, Philippe Mathieu-Daudé <philmd@linaro.org> wrote:
>
> ARM CPUs fetch instructions in little-endian.
>
> smpboot[] encoded instructions are written in little-endian.
> This is fine on little-endian host, but on big-endian ones
> the smpboot[] array ends swapped. Use the const_le32()
> macro so the instructions are always in little-endian in the
> smpboot[] array.
>
> Fixes: 9bb6d14081 ("aspeed: Add boot stub for smp booting")
> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
> ---
>  hw/arm/aspeed.c | 28 ++++++++++++++--------------
>  1 file changed, 14 insertions(+), 14 deletions(-)
>
> diff --git a/hw/arm/aspeed.c b/hw/arm/aspeed.c
> index 55f114ef72..adff9a0d73 100644
> --- a/hw/arm/aspeed.c
> +++ b/hw/arm/aspeed.c
> @@ -194,22 +194,22 @@ static void aspeed_write_smpboot(ARMCPU *cpu,
>           * r1 = AST_SMP_MBOX_FIELD_ENTRY
>           * r0 = AST_SMP_MBOX_FIELD_GOSIGN
>           */
> -        0xee100fb0,  /* mrc     p15, 0, r0, c0, c0, 5 */
> -        0xe21000ff,  /* ands    r0, r0, #255          */
> -        0xe59f201c,  /* ldr     r2, [pc, #28]         */
> -        0xe1822000,  /* orr     r2, r2, r0            */
> +        const_le32(0xee100fb0),     /* mrc     p15, 0, r0, c0, c0, 5 */
> +        const_le32(0xe21000ff),     /* ands    r0, r0, #255          */
> +        const_le32(0xe59f201c),     /* ldr     r2, [pc, #28]         */
> +        const_le32(0xe1822000),     /* orr     r2, r2, r0            */
>
> -        0xe59f1018,  /* ldr     r1, [pc, #24]         */
> -        0xe59f0018,  /* ldr     r0, [pc, #24]         */
> +        const_le32(0xe59f1018),     /* ldr     r1, [pc, #24]         */
> +        const_le32(0xe59f0018),     /* ldr     r0, [pc, #24]         */
>
> -        0xe320f002,  /* wfe                           */
> -        0xe5904000,  /* ldr     r4, [r0]              */
> -        0xe1520004,  /* cmp     r2, r4                */
> -        0x1afffffb,  /* bne     <wfe>                 */
> -        0xe591f000,  /* ldr     pc, [r1]              */
> -        AST_SMP_MBOX_GOSIGN,
> -        AST_SMP_MBOX_FIELD_ENTRY,
> -        AST_SMP_MBOX_FIELD_GOSIGN,
> +        const_le32(0xe320f002),     /* wfe                           */
> +        const_le32(0xe5904000),     /* ldr     r4, [r0]              */
> +        const_le32(0xe1520004),     /* cmp     r2, r4                */
> +        const_le32(0x1afffffb),     /* bne     <wfe>                 */
> +        const_le32(0xe591f000),     /* ldr     pc, [r1]              */
> +        const_le32(AST_SMP_MBOX_GOSIGN),
> +        const_le32(AST_SMP_MBOX_FIELD_ENTRY),
> +        const_le32(AST_SMP_MBOX_FIELD_GOSIGN)
>      };
>
>      rom_add_blob_fixed("aspeed.smpboot", poll_mailbox_ready,

Can we use the write_bootloader() function, which handles the
endianness question correctly and is how other boards do the
"put a little lump of code into the guest" job ?

thanks
-- PMM



reply via email to

[Prev in Thread] Current Thread [Next in Thread]