Hi Simon,

I want to compress a FPGA Image on the fly via binman but this doesn't work. I have add a bintool implementation for gzip, add gzip support to comp_util.py and set `compress` and `compression` property in the binman node of the u-boot dtsi:

fpga-2cg {
        compatible = "u-boot,fpga-legacy";
        description = "FPGA";
        type = "fpga";
        compression = "gzip";
        load = <(CONFIG_SYS_TEXT_BASE + 0x4000000)>;

        blob-ext {
                compress = "gzip";
                filename = "2cg.bit.bin";
        };
};

It works if I remove the `compress` property and use a `2cg.bit.bin.gz` or remove the header for gzip [1].

Regarding the code binman add the compressed size as first word to the data [2]. The code in spl-fit.c doesn't remove the size [3]. Why is this size added?

[1] https://source.denx.de/u-boot/u-boot/-/blob/master/tools/binman/comp_util.py#L44 [2] https://source.denx.de/u-boot/u-boot/-/blob/master/tools/binman/entry.py#L1065 [3] https://source.denx.de/u-boot/u-boot/-/blob/master/common/spl/spl_fit.c#L334

Regards
  Stefan

Reply via email to