[PATCH 0/3] cmd: ufetch improvements

[PATCH 0/3] cmd: ufetch improvements

[J. Neuschäfer via B4 Relay]

Following the recent addition of the "ufetch" command, this patchset
improves it in a few ways.

Signed-off-by: J. Neuschäfer <j.ne@posteo.net>
---
J. Neuschäfer (3):
      cmd: ufetch: Fix type mismatch on 32-bit
      cmd: Allow building ufetch without CONFIG_BLK
      cmd: ufetch: Show CPU architecture under "CPU"

 cmd/Kconfig  |  1 -
 cmd/ufetch.c | 14 +++++++++-----
 2 files changed, 9 insertions(+), 6 deletions(-)
---
base-commit: e5c1c2c6d9da79d874015f6d33fa2c7a34899733
change-id: 20241203-ufetch-326e771b26ad

Best regards,
-- 
J. Neuschäfer <j.ne@posteo.net>

[PATCH 1/3] cmd: ufetch: Fix type mismatch on 32-bit

[J. Neuschäfer via B4 Relay]

From: "J. Neuschäfer" <j.ne@posteo.net>

On 32-bit architectures, LAST_LINE (_LAST_LINE - 1UL) is 64 bits long,
but size_t (from ARRAY_SIZE(...)) is 32 bits. This results in a warning
because the max() macro expects the same type on both sides:

cmd/ufetch.c: In function ‘do_ufetch’:
include/linux/kernel.h:179:24: warning: comparison of distinct pointer types lacks a cast [-Wcompare-distinct-pointer-types]
  179 |         (void) (&_max1 == &_max2);              \
      |                        ^~
cmd/ufetch.c:92:25: note: in expansion of macro ‘max’
   92 |         int num_lines = max(LAST_LINE + 1, ARRAY_SIZE(logo_lines));
      |                         ^~~

Fix this by casting LAST_LINE to size_t.

Signed-off-by: J. Neuschäfer <j.ne@posteo.net>
---
 cmd/ufetch.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/cmd/ufetch.c b/cmd/ufetch.c
index 0b825d7e8c75f3b18933d3e3f77e5f40f2c7b658..5f3ef847b268dc384271fc6774720e5fd2337157 100644
--- a/cmd/ufetch.c
+++ b/cmd/ufetch.c
@@ -89,7 +89,7 @@ enum output_lines {
 static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int argc,
 		     char *const argv[])
 {
-	int num_lines = max(LAST_LINE + 1, ARRAY_SIZE(logo_lines));
+	int num_lines = max((size_t)LAST_LINE + 1, ARRAY_SIZE(logo_lines));
 	const char *model, *compatible;
 	char *ipaddr;
 	int n_cmds, n_cpus = 0, ret, compatlen;

-- 
2.45.2

Re: [PATCH 1/3] cmd: ufetch: Fix type mismatch on 32-bit

[Caleb Connolly]

Hi J,

Thanks for the patch!

On 05/12/2024 19:35, J. Neuschäfer via B4 Relay wrote:
> From: "J. Neuschäfer" <j.ne@posteo.net>
> 
> On 32-bit architectures, LAST_LINE (_LAST_LINE - 1UL) is 64 bits long,
> but size_t (from ARRAY_SIZE(...)) is 32 bits. This results in a warning
> because the max() macro expects the same type on both sides:
> 
> cmd/ufetch.c: In function ‘do_ufetch’:
> include/linux/kernel.h:179:24: warning: comparison of distinct pointer types lacks a cast [-Wcompare-distinct-pointer-types]
>   179 |         (void) (&_max1 == &_max2);              \
>       |                        ^~
> cmd/ufetch.c:92:25: note: in expansion of macro ‘max’
>    92 |         int num_lines = max(LAST_LINE + 1, ARRAY_SIZE(logo_lines));
>       |                         ^~~
> 
> Fix this by casting LAST_LINE to size_t.
> 
> Signed-off-by: J. Neuschäfer <j.ne@posteo.net>

Reviewed-by: Caleb Connolly <caleb.connolly@linaro.org>
> ---
>  cmd/ufetch.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/cmd/ufetch.c b/cmd/ufetch.c
> index 0b825d7e8c75f3b18933d3e3f77e5f40f2c7b658..5f3ef847b268dc384271fc6774720e5fd2337157 100644
> --- a/cmd/ufetch.c
> +++ b/cmd/ufetch.c
> @@ -89,7 +89,7 @@ enum output_lines {
>  static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int argc,
>  		     char *const argv[])
>  {
> -	int num_lines = max(LAST_LINE + 1, ARRAY_SIZE(logo_lines));
> +	int num_lines = max((size_t)LAST_LINE + 1, ARRAY_SIZE(logo_lines));
>  	const char *model, *compatible;
>  	char *ipaddr;
>  	int n_cmds, n_cpus = 0, ret, compatlen;
> 

-- 
// Caleb (they/them)

[PATCH 2/3] cmd: Allow building ufetch without CONFIG_BLK

[J. Neuschäfer via B4 Relay]

From: "J. Neuschäfer" <j.ne@posteo.net>

The ufetch command is still quite useful on systems without block
device support; remove the CONFIG_BLK dependency and make sure the code
compiles/works with and without CONFIG_BLK.

Signed-off-by: J. Neuschäfer <j.ne@posteo.net>
---
 cmd/Kconfig  |  1 -
 cmd/ufetch.c | 10 +++++++---
 2 files changed, 7 insertions(+), 4 deletions(-)

diff --git a/cmd/Kconfig b/cmd/Kconfig
index 4936a70f3ef16ddb093ceafa12d011ca1b89e95c..547fd2a91f7883e2ae5982897ec93c4483d67852 100644
--- a/cmd/Kconfig
+++ b/cmd/Kconfig
@@ -178,7 +178,6 @@ config CMD_CPU
 
 config CMD_UFETCH
 	bool "U-Boot fetch"
-	depends on BLK
 	help
 	  Fetch utility for U-Boot (akin to neofetch). Prints information
 	  about U-Boot and the board it is running on in a pleasing format.
diff --git a/cmd/ufetch.c b/cmd/ufetch.c
index 5f3ef847b268dc384271fc6774720e5fd2337157..7aed0b447bda104b837d37c6adcbb21b80237aba 100644
--- a/cmd/ufetch.c
+++ b/cmd/ufetch.c
@@ -92,11 +92,9 @@ static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int argc,
 	int num_lines = max((size_t)LAST_LINE + 1, ARRAY_SIZE(logo_lines));
 	const char *model, *compatible;
 	char *ipaddr;
-	int n_cmds, n_cpus = 0, ret, compatlen;
+	int n_cmds, n_cpus = 0, compatlen;
 	size_t size;
 	ofnode np;
-	struct udevice *dev;
-	struct blk_desc *desc;
 	bool skip_ascii = false;
 
 	if (argc > 1 && strcmp(argv[1], "-n") == 0) {
@@ -200,6 +198,11 @@ static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int argc,
 			break;
 		case STORAGE:
 		default:
+#ifdef CONFIG_BLK
+			struct udevice *dev;
+			struct blk_desc *desc;
+			int ret;
+
 			ret = uclass_find_device_by_seq(UCLASS_BLK, line - STORAGE, &dev);
 			if (!ret && dev) {
 				desc = dev_get_uclass_plat(dev);
@@ -213,6 +216,7 @@ static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int argc,
 			} else if (ret == -ENODEV && (skip_ascii || line > ARRAY_SIZE(logo_lines))) {
 				break;
 			}
+#endif
 			printf("\n");
 		}
 	}

-- 
2.45.2

Re: [PATCH 2/3] cmd: Allow building ufetch without CONFIG_BLK

[Caleb Connolly]

On 05/12/2024 19:35, J. Neuschäfer via B4 Relay wrote:
> From: "J. Neuschäfer" <j.ne@posteo.net>
> 
> The ufetch command is still quite useful on systems without block
> device support; remove the CONFIG_BLK dependency and make sure the code
> compiles/works with and without CONFIG_BLK.
> 
> Signed-off-by: J. Neuschäfer <j.ne@posteo.net>

Small nit below, but with that:

Reviewed-by: Caleb Connolly <caleb.connolly@linaro.org>
> ---
>  cmd/Kconfig  |  1 -
>  cmd/ufetch.c | 10 +++++++---
>  2 files changed, 7 insertions(+), 4 deletions(-)
> 
> diff --git a/cmd/Kconfig b/cmd/Kconfig
> index 4936a70f3ef16ddb093ceafa12d011ca1b89e95c..547fd2a91f7883e2ae5982897ec93c4483d67852 100644
> --- a/cmd/Kconfig
> +++ b/cmd/Kconfig
> @@ -178,7 +178,6 @@ config CMD_CPU
>  
>  config CMD_UFETCH
>  	bool "U-Boot fetch"
> -	depends on BLK
>  	help
>  	  Fetch utility for U-Boot (akin to neofetch). Prints information
>  	  about U-Boot and the board it is running on in a pleasing format.
> diff --git a/cmd/ufetch.c b/cmd/ufetch.c
> index 5f3ef847b268dc384271fc6774720e5fd2337157..7aed0b447bda104b837d37c6adcbb21b80237aba 100644
> --- a/cmd/ufetch.c
> +++ b/cmd/ufetch.c
> @@ -92,11 +92,9 @@ static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int argc,
>  	int num_lines = max((size_t)LAST_LINE + 1, ARRAY_SIZE(logo_lines));
>  	const char *model, *compatible;
>  	char *ipaddr;
> -	int n_cmds, n_cpus = 0, ret, compatlen;
> +	int n_cmds, n_cpus = 0, compatlen;
>  	size_t size;
>  	ofnode np;
> -	struct udevice *dev;
> -	struct blk_desc *desc;
>  	bool skip_ascii = false;
>  
>  	if (argc > 1 && strcmp(argv[1], "-n") == 0) {
> @@ -200,6 +198,11 @@ static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int argc,
>  			break;
>  		case STORAGE:
>  		default:
> +#ifdef CONFIG_BLK
> +			struct udevice *dev;

At least without the #ifdef, I see a warning "A label followed by a
declaration is a C23 extension". So to be on the safe side please put
this section in a block:

		default: {

...

		}

Kind regards,

> +			struct blk_desc *desc;
> +			int ret;
> +
>  			ret = uclass_find_device_by_seq(UCLASS_BLK, line - STORAGE, &dev);
>  			if (!ret && dev) {
>  				desc = dev_get_uclass_plat(dev);
> @@ -213,6 +216,7 @@ static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int argc,
>  			} else if (ret == -ENODEV && (skip_ascii || line > ARRAY_SIZE(logo_lines))) {
>  				break;
>  			}
> +#endif
>  			printf("\n");
>  		}
>  	}
> 

-- 
// Caleb (they/them)

Re: [PATCH 2/3] cmd: Allow building ufetch without CONFIG_BLK

[J. Neuschäfer]

On Mon, Dec 09, 2024 at 04:04:50PM +0100, Caleb Connolly wrote:
> On 05/12/2024 19:35, J. Neuschäfer via B4 Relay wrote:
> > From: "J. Neuschäfer" <j.ne@posteo.net>
> > 
> > The ufetch command is still quite useful on systems without block
> > device support; remove the CONFIG_BLK dependency and make sure the code
> > compiles/works with and without CONFIG_BLK.
> > 
> > Signed-off-by: J. Neuschäfer <j.ne@posteo.net>
> 
> Small nit below, but with that:
> 
> Reviewed-by: Caleb Connolly <caleb.connolly@linaro.org>
> > ---
[...]
> >  		default:
> > +#ifdef CONFIG_BLK
> > +			struct udevice *dev;
> 
> At least without the #ifdef, I see a warning "A label followed by a
> declaration is a C23 extension".

Ah, I wondered about this. Apparently I have a compiler that uses C23 as
the default, so I didn't see the warning. I'll fix it in the next revision.

> So to be on the safe side please put this section in a block:
> 
> 		default: {

An alternative that I prefer because it's a bit less intrusive with
regards to indentation, is to use a semicolon:

		default:;
			/* code with the same indentation as usual */
			/* and no closing brace to remember */

What do you think?


-- jn

Re: [PATCH 2/3] cmd: Allow building ufetch without CONFIG_BLK

[Caleb Connolly]

On 09/12/2024 17:51, J. Neuschäfer wrote:
> On Mon, Dec 09, 2024 at 04:04:50PM +0100, Caleb Connolly wrote:
>> On 05/12/2024 19:35, J. Neuschäfer via B4 Relay wrote:
>>> From: "J. Neuschäfer" <j.ne@posteo.net>
>>>
>>> The ufetch command is still quite useful on systems without block
>>> device support; remove the CONFIG_BLK dependency and make sure the code
>>> compiles/works with and without CONFIG_BLK.
>>>
>>> Signed-off-by: J. Neuschäfer <j.ne@posteo.net>
>>
>> Small nit below, but with that:
>>
>> Reviewed-by: Caleb Connolly <caleb.connolly@linaro.org>
>>> ---
> [...]
>>>  		default:
>>> +#ifdef CONFIG_BLK
>>> +			struct udevice *dev;
>>
>> At least without the #ifdef, I see a warning "A label followed by a
>> declaration is a C23 extension".
> 
> Ah, I wondered about this. Apparently I have a compiler that uses C23 as
> the default, so I didn't see the warning. I'll fix it in the next revision.
> 
>> So to be on the safe side please put this section in a block:
>>
>> 		default: {
> 
> An alternative that I prefer because it's a bit less intrusive with
> regards to indentation, is to use a semicolon:
> 
> 		default:;
> 			/* code with the same indentation as usual */
> 			/* and no closing brace to remember */

Hmm, haven't seen that one before. fwiw I think you can leave the
indentation as-is and just add the brackets.

Kind regards,
> 
> What do you think?
> 
> 
> -- jn

-- 
// Caleb (they/them)

[PATCH 3/3] cmd: ufetch: Show CPU architecture under "CPU"

[J. Neuschäfer via B4 Relay]

From: "J. Neuschäfer" <j.ne@posteo.net>

When looking at ufetch output it isn't immediately obvious which CPU
architecture the presented board has. This patch therefore adds the CPU
architecture string (for example "powerpc") to the "CPU:" line.

Signed-off-by: J. Neuschäfer <j.ne@posteo.net>
---
 cmd/ufetch.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/cmd/ufetch.c b/cmd/ufetch.c
index 7aed0b447bda104b837d37c6adcbb21b80237aba..0ee50bfb606aa2223072c63cd46b3a70ef5449fd 100644
--- a/cmd/ufetch.c
+++ b/cmd/ufetch.c
@@ -188,7 +188,7 @@ static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int argc,
 				if (ofnode_name_eq(np, "cpu"))
 					n_cpus++;
 			}
-			printf("CPU:" RESET " %d (1 in use)\n", n_cpus);
+			printf("CPU: " RESET CONFIG_SYS_ARCH ", %d (1 in use)\n", n_cpus);
 			break;
 		case MEMORY:
 			for (int j = 0; j < CONFIG_NR_DRAM_BANKS && gd->bd->bi_dram[j].size; j++)

-- 
2.45.2

Re: [PATCH 3/3] cmd: ufetch: Show CPU architecture under "CPU"

[Caleb Connolly]

Thanks for the patch!

On 05/12/2024 19:35, J. Neuschäfer via B4 Relay wrote:
> From: "J. Neuschäfer" <j.ne@posteo.net>
> 
> When looking at ufetch output it isn't immediately obvious which CPU
> architecture the presented board has. This patch therefore adds the CPU
> architecture string (for example "powerpc") to the "CPU:" line.

It would be nice to have the proper fancy name, but this is an
improvement nonetheless.
> 
> Signed-off-by: J. Neuschäfer <j.ne@posteo.net>
> ---
>  cmd/ufetch.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/cmd/ufetch.c b/cmd/ufetch.c
> index 7aed0b447bda104b837d37c6adcbb21b80237aba..0ee50bfb606aa2223072c63cd46b3a70ef5449fd 100644
> --- a/cmd/ufetch.c
> +++ b/cmd/ufetch.c
> @@ -188,7 +188,7 @@ static int do_ufetch(struct cmd_tbl *cmdtp, int flag, int argc,
>  				if (ofnode_name_eq(np, "cpu"))
>  					n_cpus++;
>  			}
> -			printf("CPU:" RESET " %d (1 in use)\n", n_cpus);
> +			printf("CPU: " RESET CONFIG_SYS_ARCH ", %d (1 in use)\n", n_cpus);

This will read like

CPU: arm 4 (1 in use)

which is a bit hard to parse at a glance. How about

CPU: arm [4 cores]

or some other separator between SYS_ARCH and the core count?

Kind regards,
>  			break;
>  		case MEMORY:
>  			for (int j = 0; j < CONFIG_NR_DRAM_BANKS && gd->bd->bi_dram[j].size; j++)
> 

-- 
// Caleb (they/them)

Re: [PATCH 3/3] cmd: ufetch: Show CPU architecture under "CPU"

[J. Neuschäfer]

On Mon, Dec 09, 2024 at 04:09:30PM +0100, Caleb Connolly wrote:
> Thanks for the patch!
> 
> On 05/12/2024 19:35, J. Neuschäfer via B4 Relay wrote:
> > From: "J. Neuschäfer" <j.ne@posteo.net>
> > 
> > When looking at ufetch output it isn't immediately obvious which CPU
> > architecture the presented board has. This patch therefore adds the CPU
> > architecture string (for example "powerpc") to the "CPU:" line.
> 
> It would be nice to have the proper fancy name, but this is an
> improvement nonetheless.

True

> > 
> > Signed-off-by: J. Neuschäfer <j.ne@posteo.net>
> > ---
[...]
> > -			printf("CPU:" RESET " %d (1 in use)\n", n_cpus);
> > +			printf("CPU: " RESET CONFIG_SYS_ARCH ", %d (1 in use)\n", n_cpus);
> 
> This will read like
> 
> CPU: arm 4 (1 in use)
> 
> which is a bit hard to parse at a glance. How about
> 
> CPU: arm [4 cores]
> 
> or some other separator between SYS_ARCH and the core count?

Ah yup, good idea. I also struggled with reading this line as it was;
"4 cores" or similar makes it clear what the number means.


-- jn