Re: [PATCH v3 10/17] hw/arm/allwinner-h3: add Boot ROM support

["Philippe Mathieu-Daudé" <philmd@redhat.com>]

On 1/15/20 12:10 AM, Niek Linnenbank wrote:
> On Tue, Jan 14, 2020 at 12:28 AM Philippe Mathieu-Daudé 
> <philmd@redhat.com <mailto:philmd@redhat.com>> wrote:
> 
>     On 1/8/20 9:00 PM, Niek Linnenbank wrote:
>      > A real Allwinner H3 SoC contains a Boot ROM which is the
>      > first code that runs right after the SoC is powered on.
>      > The Boot ROM is responsible for loading user code (e.g. a bootloader)
>      > from any of the supported external devices and writing the downloaded
>      > code to internal SRAM. After loading the SoC begins executing the
>     code
>      > written to SRAM. This commits adds emulation of the Boot ROM firmware
>      > setup functionality by loading user code from SD card.
>      >
>      > Signed-off-by: Niek Linnenbank <nieklinnenbank@gmail.com
>     <mailto:nieklinnenbank@gmail.com>>
>      > ---
>      >   include/hw/arm/allwinner-h3.h | 23 +++++++++++++++++++++++
>      >   hw/arm/allwinner-h3.c         | 28 ++++++++++++++++++++++++++++
>      >   hw/arm/orangepi.c             |  3 +++
>      >   3 files changed, 54 insertions(+)
>      >
>      > diff --git a/include/hw/arm/allwinner-h3.h
>     b/include/hw/arm/allwinner-h3.h
>      > index 5d74cca28e..4b66227ac4 100644
>      > --- a/include/hw/arm/allwinner-h3.h
>      > +++ b/include/hw/arm/allwinner-h3.h
>      > @@ -50,6 +50,7 @@
>      >   #include "hw/sd/allwinner-sdhost.h"
>      >   #include "hw/net/allwinner-sun8i-emac.h"
>      >   #include "target/arm/cpu.h"
>      > +#include "sysemu/block-backend.h"
>      >
>      >   /**
>      >    * Allwinner H3 device list
>      > @@ -130,4 +131,26 @@ typedef struct AwH3State {
>      >       MemoryRegion sram_c;
>      >   } AwH3State;
>      >
>      > +/**
>      > + * Emulate Boot ROM firmware setup functionality.
>      > + *
>      > + * A real Allwinner H3 SoC contains a Boot ROM
>      > + * which is the first code that runs right after
>      > + * the SoC is powered on. The Boot ROM is responsible
>      > + * for loading user code (e.g. a bootloader) from any
>      > + * of the supported external devices and writing the
>      > + * downloaded code to internal SRAM. After loading the SoC
>      > + * begins executing the code written to SRAM.
>      > + *
>      > + * This function emulates the Boot ROM by copying 32 KiB
>      > + * of data from the given block device and writes it to
>      > + * the start of the first internal SRAM memory.
>      > + *
>      > + * @s: Allwinner H3 state object pointer
>      > + * @blk: Block backend device object pointer
>      > + * @errp: Error object pointer for raising errors
>      > + */
>      > +void allwinner_h3_bootrom_setup(AwH3State *s, BlockBackend *blk,
>      > +                                Error **errp);
>      > +
>      >   #endif /* HW_ARM_ALLWINNER_H3_H */
>      > diff --git a/hw/arm/allwinner-h3.c b/hw/arm/allwinner-h3.c
>      > index e692432b4e..e7b768ad5b 100644
>      > --- a/hw/arm/allwinner-h3.c
>      > +++ b/hw/arm/allwinner-h3.c
>      > @@ -27,6 +27,7 @@
>      >   #include "hw/char/serial.h"
>      >   #include "hw/misc/unimp.h"
>      >   #include "hw/usb/hcd-ehci.h"
>      > +#include "hw/loader.h"
>      >   #include "sysemu/sysemu.h"
>      >   #include "hw/arm/allwinner-h3.h"
>      >
>      > @@ -168,6 +169,33 @@ enum {
>      >       AW_H3_GIC_NUM_SPI       = 128
>      >   };
>      >
>      > +void allwinner_h3_bootrom_setup(AwH3State *s, BlockBackend *blk,
>     Error **errp)
>      > +{
>      > +    uint8_t *buffer;
>      > +    int64_t rom_size = 32 * KiB;
> 
>     Why restrict to 32K? The A1 SRAM is 64K.
> 
> 
> The reason is that the actual Boot ROM on the H3 also uses 32K:
> https://linux-sunxi.org/BROM
> 
> See the 'U-Boot SPL Limitations' table at the end of the page.
> 
> You can see the comment in the table there regarding the 32 KiB:
>    "Sizes larger than 32 KiB are rejected by the BROM. Exactly 32 KiB is 
> fine, as verified by writing a special pattern at the end of the SPL and 
> checking it in the SRAM."
> Probably it would not harm to increase it to the full size of the SRAM, 
> but I tried to model
> the behavior as close to the real hardware as possible.

OK, then please document this difference in the commit description - 
such "While the A1 SRAM is 64K, we limit to 32K because ..." - and add a 
reference to https://linux-sunxi.org/BROM#U-Boot_SPL_limitations