On Thu, Feb 26, 2026 at 08:15:14AM +0000, Dorde Stojicevic wrote: > Hi all, > > yup, I completely understand and agree with you. > > One way would be to detect the characteristic string Lynx has in the > beginning of the FIT image " eLynxSecure" : > > mkimage: > FIT description: LynxSecure 2025.10.0-39844b80e0 SRP (aarch64) > Created: Thu Feb 19 11:44:07 2026 > Image 0 (kernel-1) > Description: LynxSecure 2025.10.0-39844b80e0 SRP (aarch64) > Created: Thu Feb 19 11:44:07 2026 > Type: Kernel Image > Compression: uncompressed > Data Size: > Architecture: AArch64 > OS: Linux > Load Address: > Entry Point: > Image 1 (fdt-1) > Description: Flattened Device Tree blob for LynxSecure > 2025.10.0-39844b80e0 SRP (aarch64) > Created: Thu Feb 19 11:44:07 2026 > Type: Flat Device Tree > Compression: uncompressed > Data Size: > Architecture: AArch64 > Default Configuration: 'conf-1' > Configuration 0 (conf-1) > Description: LynxSecure 2025.10.0-39844b80e0 SRP (aarch64) > Kernel: kernel-1 > FDT: fdt-1 > > xxd -l 128 > 00000000: d00d feed 1dea 3bd5 0000 0030 1dea 3b64 ......;....0..;d > 00000010: 0000 0030 0000 0011 0000 0010 0000 0000 ...0............ > 00000020: 0000 0071 1dea 3b34 0000 0000 0000 0000 ...q..;4........ > 00000030: 0000 0001 0000 0000 0000 0003 0000 002e ................ > 00000040: 0000 0065 4c79 6e78 5365 6375 7265 2032 ...eLynxSecure 2 > 00000050: 3032 352e 3130 2e30 2d33 3938 3434 6238 025.10.0-39844b8 > 00000060: 3065 3020 5352 5020 2861 6172 6368 3634 0e0 SRP (aarch64 > 00000070: 2900 0000 0000 0003 0000 0004 0000 0056 )..............V > > Thing is I am currently pretty busy with other stuff, so wont be able to > commit myself to it at the moment. > Maybe you can work out a proper solution, based on info I provided. > To my understanding, shouldn’t actually be that complex, e.g. just add check > for eLynxSecure in the code location where my patch is. > But may be I don’t see the whole picture of U-Boot and missing something. > > Let me know if I can be of help somehow
I was suggesting adding a new IH_OS type because otherwise we're going to be opening ourselves up to invalid (or malicious) images being passed in. Thanks! -- Tom
signature.asc
Description: PGP signature
