Hi Simon, On Fri, Sep 14, 2018 at 12:54 PM, Simon Glass <s...@chromium.org> wrote: > Hi Jens, > > On 3 September 2018 at 16:47, Jens Wiklander <jens.wiklan...@linaro.org> > wrote: >> With CONFIG_SANDBOX malloc a buffer in get_sector_buf() and return that >> instead of returning CONFIG_FASTBOOT_BUF_ADDR since there's no such >> buffer reserved. >> >> Signed-off-by: Jens Wiklander <jens.wiklan...@linaro.org> >> --- >> include/avb_verify.h | 9 +++++++++ >> 1 file changed, 9 insertions(+) >> >> diff --git a/include/avb_verify.h b/include/avb_verify.h >> index a532a2331aea..7b95409d247e 100644 >> --- a/include/avb_verify.h >> +++ b/include/avb_verify.h >> @@ -76,7 +76,16 @@ static inline size_t get_sector_buf_size(void) >> >> static inline void *get_sector_buf(void) >> { >> +#ifdef CONFIG_SANDBOX >> + static void *p; >> + >> + if (!p) >> + p = avb_malloc_(get_sector_buf_size()); >> + >> + return p; >> +#else >> return (void *)CONFIG_FASTBOOT_BUF_ADDR; > > Can you not define this on sandbox? We don't want sandbox to have > special-case code if we can avoid it. > > If you do define it, you should change the code to: > > return map_sysmem(CONFIG_..., size); > > which works on sandbox and other archs too.
Thanks, I'll use that. -- Jens _______________________________________________ U-Boot mailing list U-Boot@lists.denx.de https://lists.denx.de/listinfo/u-boot