KeyError: '.flash.rodata'

[danyhm]

Similar to #8 I get this error

C:\Desktop\badge\esp\esp32_image_parser>python esp32_image_parser.py create_elf new.bin -partition app0 -output app0.elf
Dumping partition 'app0' to app0_out.bin
Warning: Unexpected chip ID in image. Expected 0 but value was 9. Is this image for a different chip model?
Traceback (most recent call last):
  File "C:\Desktop\badge\esp\esp32_image_parser\esp32_image_parser.py", line 283, in <module>
    main()
  File "C:\Desktop\badge\esp\esp32_image_parser\esp32_image_parser.py", line 266, in main
    image2elf(dump_file, output_file, verbose)
  File "C:\Desktop\badge\esp\esp32_image_parser\esp32_image_parser.py", line 161, in image2elf
    size = len(section_data[name]['data'])
               ~~~~~~~~~~~~^^^^^^
KeyError: '.flash.rodata'

any ideas?

[rcorsi]

Are you using firmware from an ESP32-S3 or S2? I'm having same issue with ESP32-S3.

Output below seems to indicate that we are not dealing with an esp32

Warning: Unexpected chip ID in image. Expected 0 but value was 9. Is this image for a different chip model?

I think this repo is getting out of date with all the newer chips.

I changed this line

diff --git a/esp32_image_parser.py b/esp32_image_parser.py
index 6503cf7..ef487c4 100755
--- a/esp32_image_parser.py
+++ b/esp32_image_parser.py
@@ -38,7 +38,7 @@ def calcPhFlg(flags):
     return p_flags
 
 def image2elf(filename, output_file, verbose=False):
-    image = LoadFirmwareImage('esp32', filename)
+    image = LoadFirmwareImage('esp32s3', filename)
 
     # parse image name
     # e.g. 'image.bin' turns to 'image'

and now I get different output and chip ID warning is gone

$ python3 ./esp32_image_parser/image2elf.py firmware.bin firmware.elf
Unsure what to do with segment: BYTE_ACCESSIBLE, MEM_INTERNAL, DRAM
Unsure what to do with segment: MEM_INTERNAL, IRAM
Unsure what to do with segment: MEM_INTERNAL, IRAM
Traceback (most recent call last):
  File "/home/bobo/projects/./esp32_image_parser/image2elf.py", line 16, in <module>
    main()
    ~~~~^^
  File "/home/bobo/projects/./esp32_image_parser/image2elf.py", line 12, in main
    image2elf(args.input, args.output, args.v)
    ~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/bobo/projects/esp32_image_parser/esp32_image_parser.py", line 161, in image2elf
    size = len(section_data[name]['data'])
               ~~~~~~~~~~~~^^^^^^
KeyError: '.dram0.data'

[rcorsi]

Looking into #9 #11 and #17

[rcorsi]

I got idea to change the chip ID in #12

[rcorsi]

I think this fork might have what is needed to solve the problem https://github.com/precurse/esp32_image_parser

Look in the extra 2 commits.

[ThierryBesancon]

Hello

I encounter the same bug as @danyhm but with https://github.com/precurse/esp32_image_parser that @rcorsi told about (i was already using this fork from this repo). All i can say is the fork by no means fixes the problem.

I am using these softwares because i am a CTF player trying to solve challenges that involve ESP32 in this case. I am not sure my patch is production ready for any other use than CTF.

I patched the fork to workaround the problem that way :

--- esp32_image_parser.py	2024-11-13 02:47:33.000000000 +0100
+++ esp32_image_parser.py	2026-07-12 18:17:59.085728451 +0200
@@ -156,6 +156,9 @@
     ###########################
     print_verbose(verbose, "\nAdding program headers")
     for (name, flags) in segments.items():
+        if ( name not in section_data):
+            print_verbose(verbose, f"\nSkipping absent section {name}")
+            continue
 
         if (name == '.iram0.vectors'):
             # combine these

I just try if we have an existing section and if not i skip to the next section. In my case it works.

Thierry