Hi Alexey,
On 7/13/26 12:02 PM, Alexey Charkov wrote:
From: Jonas Karlman <[email protected]>
The v2 image format can embed up to 4 data files compared to the two
init and boot data files using the older image format.
Add support for displaying more of the image header information that
exists in the v2 image format, e.g. image load address and flag.
Example for v2 image format:
> tools/mkimage -l rk3576_idblock_v1.09.107.img
Rockchip Boot Image (v2)
Image 1: 4096 @ 0x1000
- Load address: 0x3ffc0000
Image 2: 77824 @ 0x2000
- Load address: 0x3ff81000
Image 3: 262144 @ 0x15000
Example for older image format:
> tools/mkimage -l u-boot-rockchip.bin
Rockchip RK32 (SD/MMC) Boot Image
Init Data: 20480 @ 0x800
Boot Data: 112640 @ 0x5800
Signed-off-by: Jonas Karlman <[email protected]>
Signed-off-by: Alexey Charkov <[email protected]>
---
tools/rkcommon.c | 41 +++++++++++++++++++++++++++++++----------
1 file changed, 31 insertions(+), 10 deletions(-)
diff --git a/tools/rkcommon.c b/tools/rkcommon.c
index 034896d57f80..fe0afc5629b3 100644
--- a/tools/rkcommon.c
+++ b/tools/rkcommon.c
@@ -350,8 +350,6 @@ static void rkcommon_set_header0_v2(void *buf, struct
image_tool_params *params)
uint8_t *image_ptr = NULL;
int i;
- printf("Image Type: Rockchip %s boot image\n",
- rkcommon_get_spl_hdr(params));
memset(buf, '\0', RK_INIT_OFFSET * RK_BLK_SIZE);
hdr->magic = cpu_to_le32(RK_MAGIC_V2);
hdr->boot_flag = cpu_to_le32(HASH_SHA256);
@@ -505,6 +503,29 @@ int rkcommon_verify_header(unsigned char *buf, int size,
return -ENOENT;
}
+static void rkcommon_print_header_v2(const struct header0_info_v2 *hdr)
+{
+ uint32_t val;
Scope can be reduced by moving this into the for-loop below.
+ int i;
+
+ printf("Rockchip Boot Image (v2)\n");
+
+ for (i = 0; i < le16_to_cpu(hdr->num_images); i++) {
+ printf("Image %u: %u @ 0x%x\n",
+ le32_to_cpu(hdr->images[i].counter),
+ le16_to_cpu(hdr->images[i].size) * RK_BLK_SIZE,
+ le16_to_cpu(hdr->images[i].offset) * RK_BLK_SIZE);
+
+ val = le32_to_cpu(hdr->images[i].address);
+ if (val != 0xFFFFFFFF)
+ printf("- Load address: 0x%x\n", val);
Why? Isn't it always useful info?
+
+ val = le32_to_cpu(hdr->images[i].flag);
+ if (val)
+ printf("- Flag: 0x%x\n", val);
Considering we never set the flag member, I don't think this makes sense
until we add code that sets it. I see you also print something in the
next commit that U-Boot does not set, so just make the flag printing its
own commit and justify as coming from binaries generated by boot_merger.
+ }
+}
+
void rkcommon_print_header(const void *buf, struct image_tool_params *params)
{
struct header0_info header0;
@@ -521,8 +542,7 @@ void rkcommon_print_header(const void *buf, struct
image_tool_params *params)
return;
}
- init_size = le16_to_cpu(header0_v2.images[0].size) * RK_BLK_SIZE;
- boot_size = le16_to_cpu(header0_v2.images[1].size) *
RK_BLK_SIZE;
+ rkcommon_print_header_v2(&header0_v2);
} else {
ret = rkcommon_parse_header(buf, &header0, &spl_info);
@@ -540,15 +560,16 @@ void rkcommon_print_header(const void *buf, struct image_tool_params *params)
boot_size = le16_to_cpu(header0.init_boot_size) * RK_BLK_SIZE -
init_size;
- printf("Image Type: Rockchip %s (%s) boot image\n",
- spl_info->spl_hdr,
+ printf("Rockchip %s (%s) Boot Image\n", spl_info->spl_hdr,
Please keep printing "Image Type: ", from a quick glance most other
image types also print that prefix (same remark for v2).
(image_type == IH_TYPE_RKSD) ? "SD/MMC" : "SPI");
- }
- printf("Init Data Size: %d bytes\n", init_size);
+ printf("Init Data: %d @ 0x%x\n", init_size,
+ le16_to_cpu(header0.init_offset) * RK_BLK_SIZE);
- if (boot_size != RK_MAX_BOOT_SIZE)
- printf("Boot Data Size: %d bytes\n", boot_size);
+ if (boot_size != RK_MAX_BOOT_SIZE)
+ printf("Boot Data: %d @ 0x%x\n", boot_size, init_size +
+ le16_to_cpu(header0.init_offset) * RK_BLK_SIZE);
+ }
Just split the logic for v1 into a function like you did for v2 so we're
consistent. A single commit for moving both
Do *NOT* reword/capitalize/rephrase/change what's printed for the v1.
Even though we don't specify this is stable or part of an API, I'm sure
some people out there are parsing mkimage -l's output so let's not break
them unless we have to. For v2 I guess it's fine since it'll be heavily
modified and I never was a fan of the Init Data and Boot Data (what does
this actually mean....).
Cheers,
Quentin
