Use <stdlib.h> for alloca on FreeBSD.

[iphydf]

https://www.freebsd.org/cgi/man.cgi?alloca

Some additional fixed had to be done as the Windows build started failing once I fixed the freebsd build.

---

This change is 

[robinlinden]

---

Reviewed 9 of 9 files at r1.
Review status: all files reviewed at latest revision, all discussions resolved.

---

Comments from Reviewable

[sudden6]

Reviewed 7 of 9 files at r1.
Review status: all files reviewed at latest revision, 2 unresolved discussions, all commit checks successful.

---

other/travis/toxcore-script, line 13 at r1 (raw file):

}

# Try some combinations of cmake options.

why is this removed? not needed anymore?

---

toxcore/Messenger.c, line 2261 at r1 (raw file):

            uint8_t filenumber = data[0];

#if UINT8_MAX > MAX_CONCURRENT_FILE_PIPES

What happens in the case of UINT8_MAX < MAX_CONCURRENT_FILE_PIPES? looks a bit like a hack to pass tests for me

---

Comments from Reviewable

[iphydf]

Review status: all files reviewed at latest revision, 2 unresolved discussions, all commit checks successful.

---

toxcore/Messenger.c, line 2261 at r1 (raw file):

Previously, sudden6 wrote…

What happens in the case of UINT8_MAX < MAX_CONCURRENT_FILE_PIPES? looks a bit like a hack to pass tests for me

This is not to pass tests. The condition below is always false if the cpp condition is false. filenumber is a uint8_t, so can never be larger than 255. This causes compiler warnings and I'd like to avoid those. This has nothing to do with tests.

If UINT8_MAX < MAX..., then that means we can't even index the highest filenumber. I don't think that's possible without changing the type of filenumber in a lot of places.

---

Comments from Reviewable

[iphydf]

Review status: all files reviewed at latest revision, 2 unresolved discussions, all commit checks successful.

---

other/travis/toxcore-script, line 13 at r1 (raw file):

Previously, sudden6 wrote…

why is this removed? not needed anymore?

It has never caught any bugs, and seriously slows down windows and freebsd builds.

---

Comments from Reviewable

[sudden6]

Review status: all files reviewed at latest revision, 2 unresolved discussions.

---

toxcore/Messenger.c, line 2261 at r1 (raw file):

Previously, iphydf wrote…

This is not to pass tests. The condition below is always false if the cpp condition is false. filenumber is a uint8_t, so can never be larger than 255. This causes compiler warnings and I'd like to avoid those. This has nothing to do with tests.

If UINT8_MAX < MAX..., then that means we can't even index the highest filenumber. I don't think that's possible without changing the type of filenumber in a lot of places.

Ok, can we enforce UINT8_MAX >= MAX_CONCURRENT_FILE_PIPES by a #error + comment? Would make sense against accidental changes IMO.

---

Comments from Reviewable

[sudden6]

Reviewed 1 of 9 files at r1.
Review status: all files reviewed at latest revision, all discussions resolved.

---

Comments from Reviewable

[iphydf]

Review status: all files reviewed at latest revision, 1 unresolved discussion.

---

toxcore/Messenger.c, line 2261 at r1 (raw file):

Previously, sudden6 wrote…

Ok, can we enforce UINT8_MAX >= MAX_CONCURRENT_FILE_PIPES by a #error + comment? Would make sense against accidental changes IMO.

Done.

---

Comments from Reviewable

[sudden6]

---

Reviewed 2 of 2 files at r2.
Review status: all files reviewed at latest revision, all discussions resolved.

---

Comments from Reviewable