4598: Fix console line y offsets when resizing r=def- a=Jupeyy
Hopefully fixes#4319
I can only imagine it happening bcs of resizing
## Checklist
- [x] Tested the change ingame (only tested with resizing)
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: Jupeyy <jupjopjap@gmail.com>
4594: Fix OOB access in CSnapshotDelta::UnpackDelta, cleanup r=def- a=Robyt3
The `Type` unpacked from the snapshot delta is stored in `m_SnapshotCurrent`, which is validated to be in range `0 - 0xffff`, but the arrays `m_aSnapshotDataUpdates` and `m_aSnapshotDataRate` being index with this are sized 1 too small with length `0xffff`. Hence, snapshot deltas with `Type == 0xffff` cause OOB accesses.
This is fixed by adding validation for `Type` to be in range `0 - CSnapshot::MAX_TYPE` (`0x7fff`) and changing the lengths of the arrays to `CSnapshot::MAX_TYPE + 1`. Validation for `ID` is also added to keep it in range `0 - 0xffff`.
This validation also prevents undefined behavior in subsequent `(Type << 16) | ID`.
Refactoring:
- Declare `CSnapshotDelta::UndiffItem` as `static` and add output parameter `int *pDataRate` as replacement for usage of member variables. Remove `CSnapshotDelta::m_SnapshotCurrent` that was only used for this purpose.
- Use `const` where possible.
- Merge variable declaration with first assignment. Move declarations closer to where variables are used.
- Remove dead code.
## Checklist
- [X] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [X] Considered possible null pointers and out of bounds array indexing
- [X] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: Robert Müller <robert.mueller@uni-siegen.de>
4588: Upgrade SDL2 2.0.18 -> 2.0.20 (except on Windows) r=def- a=def-
Still works on macOS
## Checklist
- [x] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: Dennis Felsing <dennis@felsin9.de>
4592: Fix crash in macOS when moving screen (fixes#4435) r=Jupeyy a=def-
Thanks to Jupeyy for the suggestion
Normally it crashes after a few seconds, tried moving window around, resizing, moving between screens like crazy and no crashes after 1-2 min.
## Checklist
- [x] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: Dennis Felsing <dennis@felsin9.de>
4582: Revert "Disable gfx_asyncrender_old on macOS (fixes#4435)" r=Jupeyy a=def-
Doesn't help fully against the crashes, just makes them a bit less
likely. Causes VSync to have even more delay. Should find an alternative
(real) fix.
This reverts commit 7d8a142768.
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [ ] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: def <dennis@felsin9.de>
4587: Fix OOB read in snapshot code reported by mmmds r=def- a=Robyt3
See https://github.com/teeworlds/teeworlds/issues/2643.
## Checklist
- [X] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [X] Considered possible null pointers and out of bounds array indexing
- [X] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: heinrich5991 <heinrich5991@gmail.com>
4586: Fail database queries faster r=def- a=Zwelf
Makes ddos situations more bearable during tournaments. Redo of #4429, but this time the tested version. I somehow mixed my two git environments (Laptop and PC) last time and some changes were missing from the PR last time (the initialization of the shutdown variable and m_count variable).
Changed the logic a bit from last time to only go into FailMode if the query fails on all database servers.
## Checklist
- [x] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [x] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: Zwelf <zwelf@strct.cc>
Doesn't help fully against the crashes, just makes them a bit less
likely. Causes VSync to have even more delay. Should find an alternative
(real) fix.
This reverts commit 7d8a142768.
4574: Always clear stars, also when entering dummy (fixes#4554) r=edg-l a=def-
otherwise they get stuck. Thanks to Skeith reproduction steps:
> Basically, to consistently reproduce the bug, you put your tee in
> freeze, then you connect your dummy right as the stars appear when
> frozen. It also happens with the stars that pop out when you shoot your
> pistol `@deen`
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [x] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: Dennis Felsing <dennis@felsin9.de>
otherwise they get stuck. Thanks to Skeith reproduction steps:
> Basically, to consistently reproduce the bug, you put your tee in
> freeze, then you connect your dummy right as the stars appear when
> frozen. It also happens with the stars that pop out when you shoot your
> pistol @deen
4571: Add CheckAtomic.cmake r=def- a=yangfl
Sometimes linking against libatomic is required for atomic ops, if
the platform doesn't support lock-free atomics.
This is a fixup for #4556, which does not work as intended.
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [ ] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: yangfl <yangfl@users.noreply.github.com>
Sometimes linking against libatomic is required for atomic ops, if
the platform doesn't support lock-free atomics.
This is a fixup for #4556, which does not work as intended.
4570: No need to create empty directory r=heinrich5991 a=def-
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [ ] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: def <dennis@felsin9.de>
4566: Don't crash when no ids are left (fixes#4565) r=heinrich5991 a=def-
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [ ] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: def <dennis@felsin9.de>
4529: Extend DDNet-Server-Launcher, support DDNet-Server from client in dmg (fixes#4441) r=heinrich5991 a=def-
The old version would just open the file opener immediately which was quite confusing for new users.
Also removed sqlite3 for mac and mysql entirely since they are not needed.
## Checklist
- [x] Tested the change ingame
- [x] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
4530: Create custom look for DMG using dmgbuild (fixes#4442) r=heinrich5991 a=def-
This removes cross-building DMG support since dmgbuild requires hdiutil, which is macOS only. Thoughts?
Nice background image by Ravie:
<img width="752" alt="Screenshot 2021-12-27 at 00 29 43" src="https://user-images.githubusercontent.com/2335377/147422479-fb278e53-5017-4176-81e8-7cb2e6c83da4.png">
Co-authored-by: def <dennis@felsin9.de>
4551: Revert "Support 32bit color depth, default to it (fixes#4549)" r=heinrich5991 a=def-
This reverts commit ee35097385.
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [ ] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
4557: Editor: Don't react to server settings shortcuts when dialog is open r=heinrich5991 a=def-
as reported by Sorah
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [x] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: def <dennis@felsin9.de>
4542: Downgrade SDL 2.0.18 -> 2.0.16 on Windows (fixes#4537) r=heinrich5991 a=def-
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [ ] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: def <dennis@felsin9.de>
4560: Support 128 players in serverbrowser r=def- a=fokkonaut
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [x] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: fokkonaut <35420825+fokkonaut@users.noreply.github.com>
4556: Add CheckAtomic.cmake r=heinrich5991 a=yangfl
Sometimes linking against libatomic is required for atomic ops, if
the platform doesn't support lock-free atomics.
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [ ] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: yangfl <yangfl@users.noreply.github.com>
4555: Fix undefined behavior in byte packing functions, add tests r=def- a=Robyt3
- Fix undefined/implementation-specific behavior due to right shift of negative number in `int_to_bytes_be` and `bytes_be_to_int`, by instead copying the bytes without using any bit arithmetic.
- Add tests for all four byte packing functions.
## Checklist
- [X] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [X] Written a unit test if it works standalone, system.c especially
- [X] Considered possible null pointers and out of bounds array indexing
- [X] Changed no physics that affect existing maps
- [X] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: Robert Müller <robert.mueller@uni-siegen.de>
4547: Fix french typo r=def- a=35niavlys
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [ ] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: Sylvain <35niavlys@gmail.com>
4550: Support 32bit color depth, default to it (fixes#4549) r=heinrich5991 a=def-
<!-- What is the motivation for the changes of this pull request -->
## Checklist
- [x] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: def <dennis@felsin9.de>
4544: Remove compatible version filter r=def- a=heinrich5991
It doesn't do anything right now, all servers advertise a compatible
version.
It might become a compatibility hazard in the future, e.g. DDNet servers
try to advertise a different version string to 0.6 than to 0.7 clients,
because the compatible version filter on 0.6 insists that it starts with
"0.6" and on 0.7 that it starts with "0.7", clearly incompatible.
## Checklist
- [ ] Tested the change ingame
- [ ] Provided screenshots if it is a visual change
- [ ] Tested in combination with possibly related configuration options
- [ ] Written a unit test if it works standalone, system.c especially
- [ ] Considered possible null pointers and out of bounds array indexing
- [ ] Changed no physics that affect existing maps
- [ ] Tested the change with [ASan+UBSan or valgrind's memcheck](https://github.com/ddnet/ddnet/#using-addresssanitizer--undefinedbehavioursanitizer-or-valgrinds-memcheck) (optional)
Co-authored-by: heinrich5991 <heinrich5991@gmail.com>
