Hi Alex,
On 22 June 2018 at 15:29, Alexander Graf agraf@suse.de wrote:
Am 22.06.2018 um 21:28 schrieb Simon Glass sjg@chromium.org:
Hi Alex,
On 22 June 2018 at 06:10, Alexander Graf agraf@suse.de wrote:
On 06/21/2018 09:45 PM, Simon Glass wrote:
Hi Alex,
On 21 June 2018 at 03:58, Alexander Graf agraf@suse.de wrote:
On 06/21/2018 04:02 AM, Simon Glass wrote:
Hi Alex,
> On 20 June 2018 at 02:51, Alexander Graf agraf@suse.de wrote: > >> On 06/20/2018 12:02 AM, Simon Glass wrote: >> >> Hi Alex, >> >>> On 18 June 2018 at 08:45, Alexander Graf agraf@suse.de wrote: >>> >>>> On 06/18/2018 04:08 PM, Simon Glass wrote: >>>> >>>> Use a starting address of 256MB which should be available. This >>>> helps >>>> to >>>> make sandbox RAM buffers pointers more recognisable. >>>> >>>> Signed-off-by: Simon Glass sjg@chromium.org >>> >>> >>> Nak, this has a non-0 chance of failing, in case something else is >>> already >>> mapped at that address. You don't want to have your CI blow up 1% of >>> the >>> time due to this. >> >> It's just a hint though. Everything will still work if it doesn't get >> this exact address. > > > I don't see what it buys us then.
These are my thoughts:
- We get an address before 4GB which is needed for grub (so far as I
can tell)
We only need that in the memory map which you want virtual (U-Boot address space) anyway. So there's no need to also have the Linux address be <4GB.
Grub cannot work without <4GB, right? I don't mind either way, but I think it that if we are picking an address, picking a smaller one is better.
Only if you expose host pointers as memory addresses. We don't anymore, and grub doesn't care whether a pointer is >4G.
We have to provide grub with pointers to memory with sandbox. These will be pointers into the sandbox RAM buffer. If this buffer is >=4GB then grub will not work, from my testing.
It works just fine for me?
OK that's strange, but perhaps I am building grub wrong. I'l ltry it again when things land.
Regards, Simon