compile warnings with gcc 11.2 in brotli

[GitMensch]

[  1%] Building C object CMakeFiles/brotli.dir/third-party/brotli/dec/decode.c.o
/home/install/rr/third-party/brotli/dec/decode.c:1954:41: warning: argument 2 of type ‘const uint8_t *’ {aka ‘const unsigned char *’} declared as a pointer [-Wvla-parameter]
 1954 |     size_t encoded_size, const uint8_t* encoded_buffer, size_t* decoded_size,
      |                          ~~~~~~~~~~~~~~~^~~~~~~~~~~~~~
In file included from /home/install/rr/third-party/brotli/dec/decode.c:7:
/home/install/rr/third-party/brotli/include/brotli/decode.h:204:19: note: previously declared as a variable length array ‘const uint8_t[*decoded_size]’ {aka ‘const unsigned char[*decoded_size]’}
  204 |     const uint8_t encoded_buffer[BROTLI_ARRAY_PARAM(encoded_size)],
      |     ~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/home/install/rr/third-party/brotli/dec/decode.c:1955:14: warning: argument 4 of type ‘uint8_t *’ {aka ‘unsigned char *’} declared as a pointer [-Wvla-parameter]
 1955 |     uint8_t* decoded_buffer) {
      |     ~~~~~~~~~^~~~~~~~~~~~~~
In file included from /home/install/rr/third-party/brotli/dec/decode.c:7:
/home/install/rr/third-party/brotli/include/brotli/decode.h:206:13: note: previously declared as a variable length array ‘uint8_t[encoded_size]’ {aka ‘unsigned char[encoded_size]’}
  206 |     uint8_t decoded_buffer[BROTLI_ARRAY_PARAM(*decoded_size)]);
      |     ~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
[  2%] Building C object CMakeFiles/brotli.dir/third-party/brotli/enc/encode.c.o
/home/install/rr/third-party/brotli/enc/encode.c:1433:20: warning: argument 5 of type ‘const uint8_t *’ {aka ‘const unsigned char *’} declared as a pointer [-Wvla-parameter]
 1433 |     const uint8_t* input_buffer, size_t* encoded_size,
      |     ~~~~~~~~~~~~~~~^~~~~~~~~~~~
In file included from /home/install/rr/third-party/brotli/enc/encode.c:9:
/home/install/rr/third-party/brotli/include/brotli/encode.h:293:19: note: previously declared as a variable length array ‘const uint8_t[*encoded_size]’ {aka ‘const unsigned char[*encoded_size]’}
  293 |     const uint8_t input_buffer[BROTLI_ARRAY_PARAM(input_size)],
      |     ~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/home/install/rr/third-party/brotli/enc/encode.c:1434:14: warning: argument 7 of type ‘uint8_t *’ {aka ‘unsigned char *’} declared as a pointer [-Wvla-parameter]
 1434 |     uint8_t* encoded_buffer) {
      |     ~~~~~~~~~^~~~~~~~~~~~~~
In file included from /home/install/rr/third-party/brotli/enc/encode.c:9:
/home/install/rr/third-party/brotli/include/brotli/encode.h:295:13: note: previously declared as a variable length array ‘uint8_t[input_size]’ {aka ‘unsigned char[input_size]’}
  295 |     uint8_t encoded_buffer[BROTLI_ARRAY_PARAM(*encoded_size)]);
      |     ~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

That happens on a fresh installed Ubuntu.

[rocallahan]

I get these. I don't think it's worth patching brotli to get rid of them.

[GitMensch]

Is there an active upstream to check / ask?

[rocallahan]

There should be

[GitMensch]

Those are fixed upstream, I'm checking for a new release, if not available I'd create a pull-request with updating it to the fixed version (which then has no release number, so would be 1.0.9.32 (which is 1.0.9, itself a re-released 1.0.8 + 32 commits).

[GitMensch]

Brotli has released version 1.1 and I'd like to come back to this.

Should this be still copied in completely (delete folder, include the unpacked version, test, commit) or isn't a git submodule preferable?
If that's preferable - should there be a copy in the rr-debugger org to be referenced or can we directly use the google git repo @ github for that

[rocallahan]

We want to keep copying it in the same way it is now. You can reference the Google Github commit the same way we currently do.