From mboxrd@z Thu Jan  1 00:00:00 1970
From: Simon Glass <sjg@chromium.org>
Date: Fri, 31 Jan 2020 11:14:20 -0700
Subject: [U-Boot] [PATCH v2 03/10] tiny-printf: Reorder code to support %p
In-Reply-To: <4c960ce5-0654-dcfb-abf6-71af79833160@ti.com>
References: <20191021232652.132164-1-sjg@chromium.org>
 <20191021232652.132164-4-sjg@chromium.org>
 <104d4667-7480-95be-1048-b8ff66adbb57@ti.com>
 <CAPnjgZ0MVTTp=-eSCwVM88SAimrEKbescwKscwJcG6vjkdtn_A@mail.gmail.com>
 <4c960ce5-0654-dcfb-abf6-71af79833160@ti.com>
Message-ID: <CAPnjgZ1hsRa2R4TPJki4OTsXn+i694Cu=B1XtLKU6_BFnUjYeg@mail.gmail.com>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de

Hi Vignesh,

On Thu, 30 Jan 2020 at 22:12, Vignesh Raghavendra <vigneshr@ti.com> wrote:
>
> Hi Simon,
>
> On 31/01/20 7:57 am, Simon Glass wrote:
> > Hi Faiz,
> >
> > On Thu, 30 Jan 2020 at 08:22, Faiz Abbas <faiz_abbas@ti.com> wrote:
> >>
> >> Hi Simon,
> >>
> >> On 22/10/19 4:56 am, Simon Glass wrote:
> >>> With a bit of code reordering we can support %p using the existing code
> >>> for ulong.
> >>>
> >>> Move the %p code up and adjust the logic accordingly.
> >>>
>
> [...]
> >>
> >> Retry time exceeded; starting again
> >> Problem booting with BOOTP
> >> SPL: failed to boot from all boot devices
> >> ### ERROR ### Please RESET the board ###
> >>
> >> Reverting this patch on the latest U-boot master fixes the issue for me.
> >>
> >> I'll look into this more deeply tomorrow. Let me know if you see
> >> something obviously wrong with the patch.
> >
> > Well one thing is that eth_env_set_enetaddr() called from the board's
> > board.c has this:
> >
> > sprintf(buf, "%pM", enetaddr);
> >
> > which is not supported with tiny-printf.
>
> That is not true. %pM is supported when SPL_NET_SUPPORT is enabled. See:
>
> https://gitlab.denx.de/u-boot/u-boot/blob/master/lib/tiny-printf.c#L183
>
> I added this specifically to support Ethernet Boot usecases on TI platforms
>
> But above commit seems to move pointer() function that formats the
> output under #ifdef DEBUG which definitely breaks %pM

OK I see. I think it is too confusing to use #ifdef DEBUG in this code.

One fix would be to change pointer() to return true if it actually
does something. I'll take a look.

This code needs tests also. Vignesh, do you feel like writing something?


>
> >
> > Before my patch it would probably produce an empty string, but now it
> > will produce garbage. Perhaps need an SPL check there.

Regards,
Simon