There are changes in the headers that make code conditionally compiled with "# if __GLIBC_HAVE_LONG_LONG" now compile unconditionally. But I haven't tracked down why there's no definition of __uint64_t.
It is also strange that the error shows as being on /usr/arm-linux-gnueabihf/include/bits/byteswap.h when that shouldn't be on the search path (and, AIUI, ~/partial-armhf-chroot/usr/include/arm-linux-gnueabihf/bits/byteswap.h should be picked up). The diff between these files is - as you might guess basically:
< #elif __GLIBC_HAVE_LONG_LONG
---
> #else
I've tried dumping the system path from CMake - but it looks right.
Logging current state of progress as it's EOD here.
There are changes in the headers that make code conditionally compiled with "# if __GLIBC_ HAVE_LONG_ LONG" now compile unconditionally. But I haven't tracked down why there's no definition of __uint64_t.
It is also strange that the error shows as being on /usr/arm- linux-gnueabihf /include/ bits/byteswap. h when that shouldn't be on the search path (and, AIUI, ~/partial- armhf-chroot/ usr/include/ arm-linux- gnueabihf/ bits/byteswap. h should be picked up). The diff between these files is - as you might guess basically:
< #elif __GLIBC_ HAVE_LONG_ LONG
---
> #else
I've tried dumping the system path from CMake - but it looks right.
Logging current state of progress as it's EOD here.