I found 19 simple issues in C++ code

[Andrey_Karpov]

I asked my colleague Egor Bredikhin to choose and check for me a small open source project, which I could investigate in between my other tasks. He sent me the log of the project 0 A.D. After the question "Why this project?" he answered: "I just played this game, a good real-time strategy". OK, then let it be 0 A.D.:).

Artcile: Good job, authors of the game 0 A.D! (19 insignificant bugs)

[asterix]

@elexis @vladislavbelov

[vladislavbelov]

1 hour ago, Andrey_Karpov said:

I asked my colleague Egor Bredikhin to choose and check for me a small open source project, which I could investigate in between my other tasks. He sent me the log of the project 0 A.D. After the question "Why this project?" he answered: "I just played this game, a good real-time strategy". OK, then let it be 0 A.D.:).

Artcile: Good job, authors of the game 0 A.D! (19 insignificant bugs)

I already answered to your post on the habr. But thanks a lot again for helping to find errors. Only few of them are known, and patches are waiting for the end of FF.

UPD. I will create tickets for missed bugs, when I'll be at home.

[elexis]

I have a feeling about N5

[macemen]

6 hours ago, elexis said:

I have a feeling about N5

That is something ASAN (address sanitizer) is very good catching at. Have you guys tried doing some runs with ASAN enabled? It does slow down things a bit (having to check every memory allocation/deallocation and access) so it might be impractical but it's worth a try.

[GunChleoc]

ASAN is lots faster than Valgrind, so it should be doable - it will slow down the program by a factor of ~2, so that's not bad at all.

[Stan`]

Now we'd need someone to try and report...

[vladislavbelov]

22 hours ago, elexis said:

I have a feeling about N5

It's not so critical as it could be, but it's still UB and should be fixed. But how it may work:

How usually (compiler depended) new[]/delete[] work:

• new[] allocates a block with a needed size + a space for size - a number of objects.
• delete[] extracts a size from the block and calls a destructor for each object and deallocates the block.

But! u8 is a basic type, not an object. So a compiler may replace new[] by new for basic types or use another trick, especially in case all actions are in the same place and the compiler knows a storing object. Also we have own pool and arena for some cases.

 

16 hours ago, macemen said:

That is something ASAN (address sanitizer) is very good catching at.

Not all, i.e. clang-802.0.42 doesn't detect it for basic types (for N5 too), only for objects, even without -O0+. Probably because of the trick above.

 

41 minutes ago, stanislas69 said:

Now we'd need someone to try and report...

We had an option for it sometime ago, if I'm not mistaken. It was called --enable-address-sanitizer. It'd good to add a regular (1 per week) build on Jenkins.

[fabio]

For reference: https://trac.wildfiregames.com/ticket/5288