Mirrors/yuzu
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
21278 commits 1 branch 0 tags 102 MiB
Commit graph

2020 commits

Author SHA1 Message Date
bunnei
dc2dd5d5a6
Merge pull request #8178 from tech-ticks/skyline-icache-fix 
hle: kernel: Invalidate entire icache in UnmapProcessMemory and UnmapCodeMemory (fixes #8174)
 2022-04-12 11:23:20 -07:00
bunnei
fd5e1e80da
Merge pull request #8157 from lat9nq/kernel-races 
kernel: Fix some data races
 2022-04-11 21:13:01 -07:00
tech-ticks
3b91d213b1 hle: kernel: Invalidate entire icache in UnmapProcessMemory and UnmapCodeMemory (fixes #8174) 2022-04-09 13:29:19 +02:00
Merry
1f275eb077 core/hle: Replace lock_guard with scoped_lock 2022-04-07 19:44:07 +01:00
Fernando S
50192eb4ad
Merge pull request #8148 from merryhime/interrupts 
dynarmic: Better interrupts
 2022-04-07 16:21:41 +02:00
lat9nq
b976cac49d k_system_control: Fix data race 
`return distribution(gen)` is a data race between a read and a write in
two threads, reported by TSan. Remove static random number generators so
they aren't using the same generator.
 2022-04-05 19:55:56 -04:00
lat9nq
983916e919 k_auto_object: Fix data race 
Change the memory order to acqure-release when we decrement the
reference count. Prevents a race with line 89 reported by TSan.
 2022-04-03 21:47:58 -04:00
lat9nq
6bcbbb29e7 k_thread: Fix data race 
TSan reports a data race between writing at cpp:1162 and reading at
h:262. Make the thread_state atomic to prevent this.
 2022-04-03 21:47:58 -04:00
lat9nq
d6a0666268 k_process: Fix data race 
TSan reported a race between thread 36 and thread 34, a read at :225 and
a write at :225 respectively. Make total_proces_running_time_ticks
atomic to avoid this race.
 2022-04-03 21:47:57 -04:00
lat9nq
5b5a1b7fa7 kernel: Fix current_process race 
TSan reported a race at :258 and :803, so make current_process an atomic
pointer.
 2022-04-03 21:47:57 -04:00
lat9nq
83b86d915a k_scheduler_lock: Fix data race 
TSan reports a race between the main thread and T37 during
IsLockedByCurrentThread and when it's set at the end of Lock(),
respectively. Set owner_thread to an atomic pointer to fix it.

Co-authored-by: bunnei <bunneidev@gmail.com>
 2022-04-03 21:47:57 -04:00
merry
f8b8af47ad dynarmic: Better interrupts 2022-04-03 16:39:48 +01:00
bunnei
bf1750664c hle: service: Add option for service interfaces to create or use the default thread. 2022-04-02 01:24:30 -04:00
bunnei
864523327f hle: kernel: Create a default thread for services that do not need their own host thread. 2022-04-02 01:24:30 -04:00
bunnei
9408100a80 hle: kernel: k_page_table: Fix implementations of LockForCodeMemory & UnlockForCodeMemory. 2022-03-26 01:49:34 -07:00
bunnei
0c75913bf2 hle: kernel: k_page_table: Implement LockMemoryAndOpen & UnlockMemory. 2022-03-26 01:46:41 -07:00
bunnei
1b3dba329a hle: kernel: svc: MapProcessMemory: Fix usage of KPageLinkedList to use physical address space. 2022-03-26 01:35:37 -07:00
bunnei
3bc0c2a625 hle: kernel: svc: CreateCodeMemory: Remove log of 'out' host pointer. 
- This does not seem terribly useful and is inconsistent with other usage.
 2022-03-26 01:34:29 -07:00
bunnei
94543f129e hle: kernel: k_code_memory: Fix usage of KPageLinkedList to use physical address space. 2022-03-26 01:33:16 -07:00
bunnei
5b667f5e52 hle: kernel: k_page_table: Implement MakeAndOpenPageGroup & MakePageGroup. 2022-03-26 01:25:52 -07:00
bunnei
c975a51ae7 hle: kernel: k_page_table: Add IsHeapPhysicalAddress method. 2022-03-26 01:23:43 -07:00
bunnei
f29410d821 hle: kernel: k_page_linked_list: Add Empty method. 2022-03-26 01:02:42 -07:00
bunnei
4e2401c11a hle: kernel: svc: UnmapProcessCodeMemory: Fix inverted alignment check. 2022-03-26 01:01:55 -07:00
bunnei
82ac66f8a4
Merge pull request #8048 from ameerj/include-purge 
general: Reduce unused includes across the project
 2022-03-21 18:03:54 -07:00
bunnei
5960d54722
Merge pull request #8040 from Morph1984/handle-table 
KHandleTable: Optimize table entry layout
 2022-03-19 23:17:37 -07:00
ameerj
ade596121b core: Reduce unused includes 2022-03-19 02:23:32 -04:00
ameerj
d618bba8a6 general: Reduce core.h includes 2022-03-18 02:13:02 -04:00
Morph
8b7d571b66 KHandleTable: Optimize table entry layout 
Since the handle type is not being used, we can reduce the amount of space each entry takes up by 4 bytes.
 2022-03-18 00:28:25 -04:00
bunnei
e95bb782f0 core: hle: kernel: init_slab_setup: Move CalculateSlabHeapGapSize to global namespace. 2022-03-14 18:14:54 -07:00
bunnei
5f3e77d93e core: hle: kernel: Allocate dummy threads on host thread storage. 
- Fixes a crash where on subsequent boots, long-lived host threads would have their dummy threads freed.
 2022-03-14 18:14:54 -07:00
bunnei
82a2463062 core: hle: kernel: Downgrade dangling objects warning to debug. 
- It is not impossible to leak kernel objects, so this is not really any issue anymore (albeit, still interesting).
 2022-03-14 18:14:54 -07:00
bunnei
f7d1929816 core: hle: kernel: Make object list container global and ensure it is reset on each emulation session. 2022-03-14 18:14:54 -07:00
bunnei
51589c5e21 core: hle: kernel: Remove server session tracking. 
- These are now allocated/managed by emulated memory, so we do not need to track and free them on shutdown.
 2022-03-14 18:14:54 -07:00
bunnei
0defac2f2a core: hle: kernel: k_process: Remove handle table finalize, reset page table. 2022-03-14 18:14:54 -07:00
bunnei
813b2ef253 core: hle: kernel: k_process: Implement thread local storage accurately. 2022-03-14 18:14:54 -07:00
bunnei
3210bc2767 core: hle: kernel: k_page_table: Add implementations of MapPages, UnmapPages, and FindFreeArea for TLS. 2022-03-14 18:14:54 -07:00
bunnei
15d9b0418f core: hle: kernel: k_slab_heap: Refresh to use guest allocations. 2022-03-14 18:14:54 -07:00
bunnei
a25cd4bb4b core: hle: kernel: Update init_slab_heap, use device memory, and add KThreadLocalPage and KPageBuffer. 
- Refreshes our slab initialization code to latest known behavior.
- Moves all guest kernel slabs into emulated device memory.
- Adds KThreadLocalPage and KPageBuffer, which we will use for accurate TLS management.
 2022-03-14 18:14:54 -07:00
bunnei
91819726b1 core: hle: kernel: k_page_buffer: Add KThreadLocalPage primitive. 2022-03-14 18:14:53 -07:00
bunnei
08434842b3 core: hle: kernel: k_page_buffer: Add KPageBuffer primitive. 2022-03-14 18:14:53 -07:00
bunnei
4a28d8cebb core: hle: kernel: k_thread: Ensure host Fiber is freed. 2022-03-14 18:14:53 -07:00
bunnei
ed67e1dd10 core: hle: kernel: k_server_session: Ensure SessionRequestManager is freed. 2022-03-14 18:14:53 -07:00
bunnei
25c0acc388 core: hle: kernel: k_thread: Update to reflect tree changes. 2022-03-14 18:14:53 -07:00
bunnei
07c9d9bdbd core: hle: kernel: Use weak_ptr where possible for SessionRequestHandler and SessionRequestManager. 2022-03-14 18:14:53 -07:00
bunnei
ce33503adf core: hle: kernel: k_memory_layout: Update kernel slab memory sizes. 2022-03-14 18:14:53 -07:00
bunnei
0f0e1c25bc core: hle: kernel: svc_types: Add ThreadLocalRegionSize. 2022-03-14 18:14:53 -07:00
bunnei
944d9186ca core: hle: kernel: k_condition_variable: Update to reflect tree changes. 2022-03-14 18:14:53 -07:00
bunnei
158c5845ab core: hle: kernel: k_address_arbiter: Update to reflect tree changes. 2022-03-14 18:14:53 -07:00
bunnei
853e58e593 hle: service: ldr: Use deterministic addresses when mapping NROs. 
- Instead of randomization, choose in-order addresses for where to map NROs into memory.
- This results in predictable behavior when debugging and consistent behavior when reproducing issues.
 2022-03-08 17:38:20 -08:00
bunnei
749f76e6fe hle: kernel: KPageTable: Improve implementations of MapCodeMemory and UnmapCodeMemory. 
- This makes these functions more accurate to the real HOS implementations.
- Fixes memory access issues in Super Smash Bros. Ultimate that occur when un/mapping NROs.
 2022-03-07 17:18:20 -08:00