Re: [PATCH kvm-unit-tests 6/8] arm/arm64: setup: Consolidate memory layout assumptions

[Alexandru Elisei <alexandru.elisei@arm.com>]

Hi Drew,

On 4/19/21 4:56 PM, Alexandru Elisei wrote:
> Hi Drew,
>
> On 4/15/21 6:25 PM, Andrew Jones wrote:
>> On Thu, Apr 15, 2021 at 05:59:19PM +0100, Alexandru Elisei wrote:
>>> Hi Drew,
>>>
>>> On 4/7/21 7:59 PM, Andrew Jones wrote:
>>>> Keep as much memory layout assumptions as possible in init::start
>>>> and a single setup function. This prepares us for calling setup()
>>>> from different start functions which have been linked with different
>>>> linker scripts. To do this, stacktop is only referenced from
>>>> init::start, making freemem_start a parameter to setup(). We also
>>>> split mem_init() into three parts, one that populates the mem regions
>>>> per the DT, one that populates the mem regions per assumptions,
>>>> and one that does the mem init. The concept of a primary region
>>>> is dropped, but we add a sanity check for the absence of memory
>>>> holes, because we don't know how to deal with them yet.
>>>>
>>>> Signed-off-by: Andrew Jones <drjones@redhat.com>
>>>> ---
>>>>  arm/cstart.S        |   4 +-
>>>>  arm/cstart64.S      |   2 +
>>>>  arm/flat.lds        |  23 ++++++
>>>>  lib/arm/asm/setup.h |   8 +--
>>>>  lib/arm/mmu.c       |   2 -
>>>>  lib/arm/setup.c     | 165 ++++++++++++++++++++++++--------------------
>>>>  6 files changed, 123 insertions(+), 81 deletions(-)
>>>>
>>>> diff --git a/arm/cstart.S b/arm/cstart.S
>>>> index 731f841695ce..14444124c43f 100644
>>>> --- a/arm/cstart.S
>>>> +++ b/arm/cstart.S
>>>> @@ -80,7 +80,9 @@ start:
>>>>  
>>>>  	/* complete setup */
>>>>  	pop	{r0-r1}
>>>> -	bl	setup
>>>> +	mov	r1, #0
>>> Doesn't that mean that for arm, the second argument to setup() will be 0 instead
>>> of stacktop?
>> The second argument is 64-bit, but we assume the upper 32 are zero.
> I didn't realize that phys_addr_t is 64bit.
>
> According to ARM IHI 0042F, page 15:
>
> "A double-word sized type is passed in two consecutive registers (e.g., r0 and r1,
> or r2 and r3). The content of the registers is as if the value had been loaded
> from memory representation with a single LDM instruction."
>
> I think r3 should be zeroed, not r1. r2 and r3 represent the 64bit value. arm is
> little endian, so the least significant 32bits will be in r2 and the most
> significant bits will be in r3. I can't figure out why r3 is zero, but moving the

Correction: r3 is zero because KVM zeroes the general purpose registers on vcpu reset.

Thanks,

Alex