[PATCH 1/2] firmware_loader: builtin: ignore 0-size firmware

[PATCH 1/2] firmware_loader: builtin: ignore 0-size firmware

[Dmitry Torokhov]

Currently, the builtin firmware loader allows 0-size firmware to be
returned successfully to drivers. This differs from all other loading
mechanisms (filesystem, sysfs fallback) which reject 0-byte files, and
forces drivers to add boilerplate size checks.

Modify firmware_request_builtin() to reject 0-size firmware. This will
also result in firmware loader falling back to other mechanisms if an
empty built-in firmware is present.

Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 drivers/base/firmware_loader/builtin/main.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/base/firmware_loader/builtin/main.c b/drivers/base/firmware_loader/builtin/main.c
index d36befebb1b9..1dcebe8e7f8e 100644
--- a/drivers/base/firmware_loader/builtin/main.c
+++ b/drivers/base/firmware_loader/builtin/main.c
@@ -53,6 +53,8 @@ bool firmware_request_builtin(struct firmware *fw, const char *name)
 
 	for (b_fw = __start_builtin_fw; b_fw != __end_builtin_fw; b_fw++) {
 		if (strcmp(name, b_fw->name) == 0) {
+			if (b_fw->size == 0)
+				return false;
 			fw->size = b_fw->size;
 			fw->data = b_fw->data;
 			return true;
-- 
2.54.0.545.g6539524ca2-goog

[PATCH 2/2] firmware_loader: builtin: fail build on empty firmware

[Dmitry Torokhov]

If an empty firmware file is supplied via CONFIG_EXTRA_FIRMWARE, the
build currently succeeds but creates an empty section. While the
firmware loader will not return such sections it is better to avoid
adding them in the first place.

Add a compile-time size check to the filechk_fwbin macro to abort the
build early if a 0-size firmware is encountered.

Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
---
 drivers/base/firmware_loader/builtin/Makefile | 32 +++++++++++--------
 1 file changed, 18 insertions(+), 14 deletions(-)

diff --git a/drivers/base/firmware_loader/builtin/Makefile b/drivers/base/firmware_loader/builtin/Makefile
index 6c067dedc01e..af5b1bbb31e4 100644
--- a/drivers/base/firmware_loader/builtin/Makefile
+++ b/drivers/base/firmware_loader/builtin/Makefile
@@ -16,20 +16,24 @@ ASM_ALIGN = $(if $(CONFIG_64BIT),3,2)
 PROGBITS  = $(if $(CONFIG_ARM),%,@)progbits
 
 filechk_fwbin = \
-	echo "/* Generated by $(src)/Makefile */"		;\
-	echo "    .section .rodata"				;\
-	echo "    .p2align 4"					;\
-	echo "_fw_$(FWSTR)_bin:"				;\
-	echo "    .incbin \"$(fwdir)/$(FWNAME)\""		;\
-	echo "_fw_end:"						;\
-	echo "    .section .rodata.str,\"aMS\",$(PROGBITS),1"	;\
-	echo "    .p2align $(ASM_ALIGN)"			;\
-	echo "_fw_$(FWSTR)_name:"				;\
-	echo "    .string \"$(FWNAME)\""			;\
-	echo "    .section .builtin_fw,\"a\",$(PROGBITS)"	;\
-	echo "    .p2align $(ASM_ALIGN)"			;\
-	echo "    $(ASM_WORD) _fw_$(FWSTR)_name"		;\
-	echo "    $(ASM_WORD) _fw_$(FWSTR)_bin"			;\
+	if [ ! -s "$(fwdir)/$(FWNAME)" ]; then				 \
+		echo "error: empty firmware: $(fwdir)/$(FWNAME)" >&2	;\
+		exit							;\
+	fi								;\
+	echo "/* Generated by $(src)/Makefile */"			;\
+	echo "    .section .rodata"					;\
+	echo "    .p2align 4"						;\
+	echo "_fw_$(FWSTR)_bin:"					;\
+	echo "    .incbin \"$(fwdir)/$(FWNAME)\""			;\
+	echo "_fw_end:"							;\
+	echo "    .section .rodata.str,\"aMS\",$(PROGBITS),1"		;\
+	echo "    .p2align $(ASM_ALIGN)"				;\
+	echo "_fw_$(FWSTR)_name:"					;\
+	echo "    .string \"$(FWNAME)\""				;\
+	echo "    .section .builtin_fw,\"a\",$(PROGBITS)"		;\
+	echo "    .p2align $(ASM_ALIGN)"				;\
+	echo "    $(ASM_WORD) _fw_$(FWSTR)_name"			;\
+	echo "    $(ASM_WORD) _fw_$(FWSTR)_bin"				;\
 	echo "    $(ASM_WORD) _fw_end - _fw_$(FWSTR)_bin"
 
 $(obj)/%.gen.S: FORCE
-- 
2.54.0.545.g6539524ca2-goog

Re: [PATCH 1/2] firmware_loader: builtin: ignore 0-size firmware

[Dmitry Torokhov]

On Sat, Apr 25, 2026 at 09:30:38PM -0700, Dmitry Torokhov wrote:
> Currently, the builtin firmware loader allows 0-size firmware to be
> returned successfully to drivers. This differs from all other loading
> mechanisms (filesystem, sysfs fallback) which reject 0-byte files, and
> forces drivers to add boilerplate size checks.
> 
> Modify firmware_request_builtin() to reject 0-size firmware. This will
> also result in firmware loader falling back to other mechanisms if an
> empty built-in firmware is present.
> 
> Signed-off-by: Dmitry Torokhov <dmitry.torokhov@gmail.com>
> ---

Gentle ping on this and the next one...

>  drivers/base/firmware_loader/builtin/main.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/drivers/base/firmware_loader/builtin/main.c b/drivers/base/firmware_loader/builtin/main.c
> index d36befebb1b9..1dcebe8e7f8e 100644
> --- a/drivers/base/firmware_loader/builtin/main.c
> +++ b/drivers/base/firmware_loader/builtin/main.c
> @@ -53,6 +53,8 @@ bool firmware_request_builtin(struct firmware *fw, const char *name)
>  
>  	for (b_fw = __start_builtin_fw; b_fw != __end_builtin_fw; b_fw++) {
>  		if (strcmp(name, b_fw->name) == 0) {
> +			if (b_fw->size == 0)
> +				return false;
>  			fw->size = b_fw->size;
>  			fw->data = b_fw->data;
>  			return true;

-- 
Dmitry

Re: [PATCH 1/2] firmware_loader: builtin: ignore 0-size firmware

[Danilo Krummrich]

On Sun Jul 12, 2026 at 12:11 AM CEST, Dmitry Torokhov wrote:
> Gentle ping on this and the next one...

Sorry for the delay, I picked this one up, but I have a question on the second
one.

Re: (subset) [PATCH 1/2] firmware_loader: builtin: ignore 0-size firmware

[Danilo Krummrich]

On Sat, 25 Apr 2026 21:30:38 -0700, Dmitry Torokhov wrote:
> [PATCH 1/2] firmware_loader: builtin: ignore 0-size firmware

Applied, thanks!

  Branch: driver-core-testing
  Tree:   git://git.kernel.org/pub/scm/linux/kernel/git/driver-core/driver-core.git

[1/2] firmware_loader: builtin: ignore 0-size firmware
      commit: d059966c0c7d

The patch will appear in the next linux-next integration (typically within 24
hours on weekdays).

The patch is in the driver-core-testing branch and will be promoted to
driver-core-next after validation.

Re: [PATCH 2/2] firmware_loader: builtin: fail build on empty firmware

[Danilo Krummrich]

On Sun Apr 26, 2026 at 6:30 AM CEST, Dmitry Torokhov wrote:
> +	if [ ! -s "$(fwdir)/$(FWNAME)" ]; then				 \
> +		echo "error: empty firmware: $(fwdir)/$(FWNAME)" >&2	;\
> +		exit							;\

Shouldn't this be 'exit 1' to actually fail the build?

> +	fi								;\

Re: [PATCH 2/2] firmware_loader: builtin: fail build on empty firmware

[Dmitry Torokhov]

On Sun, Jul 12, 2026 at 01:58:26AM +0200, Danilo Krummrich wrote:
> On Sun Apr 26, 2026 at 6:30 AM CEST, Dmitry Torokhov wrote:
> > +	if [ ! -s "$(fwdir)/$(FWNAME)" ]; then				 \
> > +		echo "error: empty firmware: $(fwdir)/$(FWNAME)" >&2	;\
> > +		exit							;\
> 
> Shouldn't this be 'exit 1' to actually fail the build?

Yes, you are absolutely right. Do you want me to resend or you can fix
it up on your end?

Thanks.

-- 
Dmitry

Re: (subset) [PATCH 1/2] firmware_loader: builtin: ignore 0-size firmware

[Danilo Krummrich]

On Sat, 25 Apr 2026 21:30:38 -0700, Dmitry Torokhov wrote:
> [PATCH 1/2] firmware_loader: builtin: ignore 0-size firmware

Applied, thanks!

  Branch: driver-core-testing
  Tree:   git://git.kernel.org/pub/scm/linux/kernel/git/driver-core/driver-core.git

[2/2] firmware_loader: builtin: fail build on empty firmware
      commit: 1160c2208fab

      [ Use 'exit 1' to properly fail the build. - Danilo ]

The patch will appear in the next linux-next integration (typically within 24
hours on weekdays).

The patch is in the driver-core-testing branch and will be promoted to
driver-core-next after validation.