Fix (crashing) error on M105 output

[bgort]

Use SERIAL_PROTOCOL for ints instead of SERIAL_PROTOCOL_F.

Also removed wayward 'address of' ampersand in setTargetHotend and setTargetBed parameters.

See #6571.

[colinrgodsey]

LGTM. Will verify shortly if you need an extra +, need to get my cube printing shoes on.

[Roxy-3D]

I think I agree the 'address of' does not serve a purpose in these function declarations. Why was that put there? If I knew the answer, I would merge this right now....

[bgort]

I have no idea why it's there, but there's no purpose for it that I can see. We're just referencing the value of the int passed in when setting the temps.

My guess is that there was some need for it in a previous version.

[Roxy-3D]

OK... We can change it back if it turns out there is a real reason for it. I think it can also be argued that the non-BugFix branch should get this change, but we should probably have a discussion about that...

I'll merge just to help get rid of some of the crashes...

[bgort]

Yeah, it's a pretty big deal. I can't imagine anyone who checks temps w/M105 can use 1.1.0?

[Roxy-3D]

Actually... part of the discussion ought to be what the default branch is... If we are NOT going to cross bug fixes over to the v1.1.0 branch... Probably the default branch should not point at it. The default branch should point at something that works for most people.

#6571 (comment)

[Grogyan]

I can't check this right now. Does this mean that host programs won't show won't show 1dp of precision in temperature read backs?

[Roxy-3D]

I believe the answer is "Yes." It maybe that switching to double is better than int.

[Tannoo]

My build from Compiled: May 3 2017, had these:

temperature.h:
static void setTargetHotend(const float& celsius, uint8_t e) {
and
static void setTargetBed(const float& celsius) {

And M105 works fine:

echo:Marlin 1.1.0-RCBugFix

echo: Last Updated: 2016-12-06 12:00 | Author: (none, default config)
Compiled: May  3 2017
echo: Machine Name: Prusa i3
echo: Free Memory: 2261  PlannerBufferBytes: 1232
echo:V36 stored settings retrieved (453 bytes)
 UBL active!
...
echo:DEBUG:ECHO
ok
echo:M105
ok T:25.0 /0.0 B:25.5 /0.0 @:0 B@:0
echo:M105
ok T:25.0 /0.0 B:25.5 /0.0 @:0 B@:0
echo:M105
ok T:25.0 /0.0 B:25.5 /0.0 @:0 B@:0
echo:M105
ok T:25.0 /0.0 B:25.5 /0.0 @:0 B@:0
echo:M105
ok T:25.0 /0.0 B:25.5 /0.0 @:0 B@:0
echo:M105
ok T:25.0 /0.0 B:25.5 /0.0 @:0 B@:0

[bgort]

Do you know which commit that is, Tannoo? Looks like it's from before 2658cc7, which would make sense.

EDIT: I don't think the ampersand was causing a problem, necessarily, but it was unnecessary and didn't make sense, given what the function does now.

[colinrgodsey]

seems to work with OctoPi so far

Recv: ok T:20 /0 B:19 /0 @:0 B@:0
Send: M105
Recv: ok T:20 /0 B:20 /0 @:0 B@:0
Send: M105
Recv: ok T:19 /0 B:20 /0 @:0 B@:0

Not a huge fan of that missing point of precision tho....

Now to try and print something!