Re: [PATCH] tools/atmelimage: add const qualifier to fix compiler warning

[Daniel Golle <daniel@makrotopia.org>]

On Wed, Feb 25, 2026 at 09:07:42AM -0600, Tom Rini wrote:
> On Wed, Feb 25, 2026 at 03:03:57PM +0000, Daniel Golle wrote:
> 
> > tools/atmelimage.c: In function ‘atmel_find_pmecc_parameter_in_token’:
> > tools/atmelimage.c:64:31: error: assignment discards ‘const’ qualifier from pointer target type [-Werror=discarded-qualifiers]
> > 64 |                         param = strstr(token, "=");
> >    |                               ^
> > cc1: all warnings being treated as errors
> > 
> > Add 'const' qualifier to variable 'param' to prevent build failing
> > due to -Werror.
> > 
> > Signed-off-by: Daniel Golle <daniel@makrotopia.org>
> > ---
> >  tools/atmelimage.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/tools/atmelimage.c b/tools/atmelimage.c
> > index 6a2d9d8feab..770f7a0de4e 100644
> > --- a/tools/atmelimage.c
> > +++ b/tools/atmelimage.c
> > @@ -55,7 +55,7 @@ static const char * const configs[] = {
> >  static int atmel_find_pmecc_parameter_in_token(const char *token)
> >  {
> >  	size_t pos;
> > -	char *param;
> > +	const char *param;
> >  
> >  	debug("token: '%s'\n", token);
> >  
> 
> How / where are you triggering this problem? This is part of mkimage and
> built on CI every time (and on macOS and Windows, in Azure). Thanks!

I'm using archlinux on my workstation which got a fairly
recent version of GCC which is used to build the host tools:
gcc (GCC) 15.2.1 20260209

When trying to use the u-boot-size-test.sh scripts you had asked me for
it failed because of passing the '-E' parameter to buildman.

Also 'dtc' fails building for the same reason and that can be fixed by
picking commit dfe057874b34c ("scripts/dtc: Update to upstream versionv
1.7.2-62-ga26ef6400bd8") from Linux.
However, that has far-reaching consequences such as the
-Wno-graph_child_address parameter no longer existing, and also dtc
became more strict and issues new warnings for many of the dts files in
U-Boot, so then the build failed because of that...

As I wrote in the other email thread, I ended up just removing the '-E'
paramter because the rabbit hole became too deep...
diff --git a/contrib/trini/u-boot-size-test.sh b/contrib/trini/u-boot-size-test.sh
index da722f5..4cef66e 100755
--- a/contrib/trini/u-boot-size-test.sh
+++ b/contrib/trini/u-boot-size-test.sh
@@ -1,7 +1,7 @@
 #!/bin/bash
 
 # Initial and constant buildman args
-ARGS="-devl -PEWM"
+ARGS="-devl -PWM"
 ALL=0
 KEEP=0