From: Ziyuan Xu <xzy.xu@rock-chips.com>
To: u-boot@lists.denx.de
Subject: [U-Boot] [PATCH] arm: re-implement proper ISB instruction for ARMv7-A
Date: Fri, 29 Jul 2016 09:06:29 +0800 [thread overview]
Message-ID: <579AAC15.2020804@rock-chips.com> (raw)
In-Reply-To: <20160729003437.GR5926@bill-the-cat>
Hi Tom,
On 2016?07?29? 08:34, Tom Rini wrote:
> On Fri, Jul 29, 2016 at 07:34:09AM +0800, Ziyuan Xu wrote:
>> Hi Tom,
>>
>> On 2016?07?29? 06:15, Tom Rini wrote:
>>> On Thu, Jul 28, 2016 at 07:03:17PM +0800, Chen-Yu Tsai wrote:
>>>> Hi,
>>>>
>>>> On Thu, Jul 28, 2016 at 6:13 PM, Ziyuan Xu <xzy.xu@rock-chips.com> wrote:
>>>>> For ARMv7-A architecture, the valid ISB instruction is asm volatile("isb").
>>>>>
>>>>> This patch fixes the U-Boot was stuck in invalidate_dcache_all() before
>>>>> booting linux kernel, which occurred on rk3288-base development board
>>>>> such as evb-rk3288, rock2-rk3288. And something output via console like:
>>>>>
>>>>> => bootz 0x2000000
>>>>> 0x02000000
>>>>> ramdisk start = 0x00000000, ramdisk end = 0x00000000
>>>>> Continuing to boot without FDT
>>>>> Initial value for argc=3
>>>>> Final value for argc=3
>>>>> using: ATAGS
>>>>>
>>>>> Starting kernel ...
>>>>>
>>>>> Linux kernel exactly the same way(see arch/arm/include/asm/barrier.h).
>>>>>
>>>>> Signed-off-by: Ziyuan Xu <xzy.xu@rock-chips.com>
>>>>> ---
>>>>>
>>>>> arch/arm/include/asm/system.h | 6 ++++--
>>>>> 1 file changed, 4 insertions(+), 2 deletions(-)
>>>>>
>>>>> diff --git a/arch/arm/include/asm/system.h b/arch/arm/include/asm/system.h
>>>>> index 2bdc0be..12d4ba0 100644
>>>>> --- a/arch/arm/include/asm/system.h
>>>>> +++ b/arch/arm/include/asm/system.h
>>>>> @@ -227,13 +227,15 @@ void __noreturn psci_system_reset(bool smc);
>>>>> */
>>>>> void save_boot_params_ret(void);
>>>>>
>>>>> -#define isb() __asm__ __volatile__ ("" : : : "memory")
>>>>> -
>>>>> #define nop() __asm__ __volatile__("mov\tr0,r0\t@ nop\n\t");
>>>>>
>>>>> #ifdef __ARM_ARCH_7A__
>>>>> +#define isb() __asm__ __volatile__ ("isb" : : : "memory")
>>>>> +
>>>>> #define wfi() __asm__ __volatile__ ("wfi" : : : "memory")
>>>>> #else
>>>>> +#define isb() __asm__ __volatile__ ("" : : : "memory")
>>>>> +
>>>>> #define wfi()
>>>>> #endif
>>>>>
>>>> arch/arm/include/asm/barriers.h already has a proper set of
>>>> ISB/DSB macros. Please consider using those instead.
>>> Please fix arch/arm/include/asm/system.h to use the macros found in
>>> barriers.h rather than have their own versions. Thanks!
>> If I understand correctly, I can change into as bellow:
>> #define isb() ISB
>> How about it?
> Well, I'd rather not have ISB and isb, just ISB, which means we might
> have to fix other places too. If that starts looking too huge, we can
> do this in steps and just do what you suggested for now and for next
> release move everything over.
As I mentioned before, arch/arm/include/asm/barriers.h defined ISB
macro. If I only want to fix the issue which I hit on rk3288 board, I
just use ISB in arch/arm/include/asm/system.h::set_cr() instead of
isb(). But isb() had been invoked in some places.
I can't verify integrallty after all revision, it involve some boards
and feature. But this does fix for rk3288, if you agree with me, could
you apply it provisionally?:-)
>
next prev parent reply other threads:[~2016-07-29 1:06 UTC|newest]
Thread overview: 16+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-07-28 10:13 [U-Boot] [PATCH] arm: re-implement proper ISB instruction for ARMv7-A Ziyuan Xu
2016-07-28 10:39 ` Alexander Graf
2016-07-28 12:03 ` Ziyuan Xu
2016-07-28 12:41 ` Sandy Patterson
2016-07-28 11:03 ` Chen-Yu Tsai
2016-07-28 11:51 ` Ziyuan Xu
2016-07-28 22:15 ` Tom Rini
2016-07-28 23:34 ` Ziyuan Xu
2016-07-29 0:34 ` Tom Rini
2016-07-29 1:06 ` Ziyuan Xu [this message]
2016-07-29 1:12 ` Tom Rini
2016-07-29 1:30 ` Ziyuan Xu
2016-07-31 3:59 ` Ziyuan Xu
2016-07-31 14:27 ` Tom Rini
2016-08-01 1:07 ` Ziyuan Xu
2016-07-28 22:16 ` Tom Rini
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=579AAC15.2020804@rock-chips.com \
--to=xzy.xu@rock-chips.com \
--cc=u-boot@lists.denx.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).