android/sound.cpp: Fix overflow warning

[softins]

Hopefully, this should fix the overflow warning that gets listed below every PR.

[hoffie]

Hrm, Android build is currently failing:

android/sound.cpp:329:24: error: no matching function for call to 'min'
    int32_t count    = std::min ( mOutBuffer.size(), to_write );
                       ^~~~~~~~
/opt/android/android-ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/../sysroot/usr/include/c++/v1/algorithm:2532:1: note: candidate template ignored: deduced conflicting types for parameter '_Tp' ('unsigned int' vs. 'int')
min(const _Tp& __a, const _Tp& __b)
^
/opt/android/android-ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/../sysroot/usr/include/c++/v1/algorithm:2543:1: note: candidate template ignored: could not match 'initializer_list<type-parameter-0-0>' against 'unsigned int'
min(initializer_list<_Tp> __t, _Compare __comp)
^
/opt/android/android-ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/../sysroot/usr/include/c++/v1/algorithm:2523:1: note: candidate function template not viable: requires 3 arguments, but 2 were provided
min(const _Tp& __a, const _Tp& __b, _Compare __comp)
^
/opt/android/android-ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/../sysroot/usr/include/c++/v1/algorithm:2552:1: note: candidate function template not viable: requires single argument '__t', but 2 arguments were provided
min(initializer_list<_Tp> __t)

[softins]

Hrm, Android build is currently failing:

Yes, I noticed, but I can’t understand why. I’m trying to set up a build environment, based on the workflow file, but haven’t succeeded yet.

[pljones]

Yes, I noticed, but I can’t understand why. I’m trying to set up a build environment, based on the workflow file, but haven’t succeeded yet.

Fork and reduce the builds done on your fork to just the failing one (and then down to just the needed bits, if that's easy enough) -- or is that what you meant?

[softins]

Yes, I noticed, but I can’t understand why. I’m trying to set up a build environment, based on the workflow file, but haven’t succeeded yet.

Fork and reduce the builds done on your fork to just the failing one (and then down to just the needed bits, if that's easy enough) -- or is that what you meant?

Well that's certainly one way to do it. I had been hoping to create a local build environment to mirror what Github does, but didn't succeed.

[ann0see]

Todo: fix error on Android. @j-santander might have an idea?

android/sound.cpp:329:24: error: no matching function for call to 'min'
100
int32_t count = std::min ( mOutBuffer.size(), to_write );
101
^~~~~~~~
102
/opt/android/android-ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/../sysroot/usr/include/c++/v1/algorithm:2532:1: note: candidate template ignored: deduced conflicting types for parameter '_Tp' ('unsigned int' vs. 'int')
103
min(const _Tp& __a, const _Tp& __b)
104
^
105
/opt/android/android-ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/../sysroot/usr/include/c++/v1/algorithm:2543:1: note: candidate template ignored: could not match 'initializer_list' against 'unsigned int'
106
min(initializer_list<_Tp> __t, _Compare __comp)
107
^
108
/opt/android/android-ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/../sysroot/usr/include/c++/v1/algorithm:2523:1: note: candidate function template not viable: requires 3 arguments, but 2 were provided
109
min(const _Tp& __a, const _Tp& __b, _Compare __comp)
110
^
111
/opt/android/android-ndk/toolchains/llvm/prebuilt/linux-x86_64/bin/../sysroot/usr/include/c++/v1/algorithm:2552:1: note: candidate function template not viable: requires single argument '__t', but 2 arguments were provided
112
min(initializer_list<_Tp> __t)
113
^

[j-santander]

Fails because you have unsigned (the return of the ring buffer size) and signed arguments (the count) in the std::min

what's this overflow warning? Tried to check the logs of recent builds but either I don't know how to check it or it wasn't there.

[j-santander]

BTW, I did a PR #1528 to replace my custom ring buffer with the CBuffer common to the rest of the project.

[pljones]

1. Is this still a problem?
2. Is the patch still clean?
3. Is the Android build still failing?

[softins]

1. Is this still a problem?

I expect so, and I believe there are others too. I have been intending to look at them all, but not got there yet.

2. Is the patch still clean?

Don't know.

3. Is the Android build still failing?

Will check when I pick it up again.

[ann0see]

Closing in favour of #2292

[hoffie]

Closing in favour of #2292

Hrm, isn't this about two different issues?
This PR touches android/sound.cpp, while #2292 touches src/buffer.cpp so it sounds different to me at first glance?

Edit: Maybe we should put even more context in PR/commit titles to simplify distinguishing. :)

[ann0see]

Android should already be fixed?

#1528 #2237

[hoffie]

Ah, right. This should be the relevant fix:
https://github.com/jamulussoftware/jamulus/pull/2237/files#diff-32ef006d0b2b2daef9b1dc9c1e43c991a5b41822a1a5ef972592c6d1b636ca48R299

I was just confused by the seemingly unrelated mention of #2292.