Re: [U-Boot] [PATCH] armv7: ls102xa: add lowlevel init

On 09/01/2016 07:32 PM, Xiaoliang Yang wrote:
Hi York,
We need enable i-cache in low level in order to improve u-boot running speed. We have skipped lowlevel_init because a silicon bug before. (it could not run some instructions in lowlevel_init until DDR init, the bug has already fixed.)
Xiaoliang,
Can you be more specific about the bug? You said it is fixed, how about existing SoCs? Will existing SoCs run into any issue with the lowlevel_init?
York

Hi york,
The silicon bug was existing in V1.0 of SoCs, we have already not supported them.
Xiaoliang Yang
-----Original Message----- From: york sun Sent: Friday, September 02, 2016 11:24 AM To: Xiaoliang Yang; u-boot@lists.denx.de Subject: Re: [PATCH] armv7: ls102xa: add lowlevel init
On 09/01/2016 07:32 PM, Xiaoliang Yang wrote:
Hi York,
We need enable i-cache in low level in order to improve u-boot running speed. We have skipped lowlevel_init because a silicon bug before. (it could not run some instructions in lowlevel_init until DDR init, the bug has already fixed.)
Xiaoliang,
Can you be more specific about the bug? You said it is fixed, how about existing SoCs? Will existing SoCs run into any issue with the lowlevel_init?
York

On 09/01/2016 08:39 PM, Xiaoliang Yang wrote:
Hi york,
The silicon bug was existing in V1.0 of SoCs, we have already not supported them.
No good. SDK may not support it, but ss long as we have users with rev 1 SoC, we need to support it in U-Boot. You can add revision check and skip lowever_init for older SoCs.
York
Xiaoliang Yang
-----Original Message----- From: york sun Sent: Friday, September 02, 2016 11:24 AM To: Xiaoliang Yang; u-boot@lists.denx.de Subject: Re: [PATCH] armv7: ls102xa: add lowlevel init
On 09/01/2016 07:32 PM, Xiaoliang Yang wrote:
Hi York,
We need enable i-cache in low level in order to improve u-boot running speed. We have skipped lowlevel_init because a silicon bug before. (it could not run some instructions in lowlevel_init until DDR init, the bug has already fixed.)
Xiaoliang,
Can you be more specific about the bug? You said it is fixed, how about existing SoCs? Will existing SoCs run into any issue with the lowlevel_init?
York
participants (2)
-
Xiaoliang Yang
-
york sun