Skip to content
Open
Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
58 commits
Select commit Hold shift + click to select a range
d2caf76
add XPU symm
zhangxiaoli73 Jun 4, 2025
059cee6
correct include
zhangxiaoli73 Jun 4, 2025
b8a5473
remove XPUGuard
zhangxiaoli73 Jun 4, 2025
3446fa6
debug
zhangxiaoli73 Jun 4, 2025
453fcf2
debug
zhangxiaoli73 Jun 9, 2025
d0f7e0b
debug
zhangxiaoli73 Jun 9, 2025
dad01a5
debug
zhangxiaoli73 Jun 9, 2025
d7553ec
debug
zhangxiaoli73 Jun 9, 2025
d7886ee
debug
zhangxiaoli73 Jun 9, 2025
002a8ec
debug
zhangxiaoli73 Jun 9, 2025
8530c74
debug
zhangxiaoli73 Jun 9, 2025
5ffc7f8
debug
zhangxiaoli73 Jun 9, 2025
e3c722a
debug
zhangxiaoli73 Jun 10, 2025
dfea229
add debug logs
zhangxiaoli73 Jun 10, 2025
2bcde21
check device type
zhangxiaoli73 Jun 12, 2025
ce6f875
debug async ops
zhangxiaoli73 Jun 12, 2025
67d9ee3
debug
zhangxiaoli73 Jun 13, 2025
eaca2bd
refine to void*
zhangxiaoli73 Jun 13, 2025
7be7d36
debug
zhangxiaoli73 Jun 13, 2025
e203d62
debug copy
zhangxiaoli73 Jun 13, 2025
8331817
debug sharded handle
zhangxiaoli73 Jun 16, 2025
0a83b56
debug
zhangxiaoli73 Jun 17, 2025
646d246
debug
zhangxiaoli73 Jun 17, 2025
fed3327
enable torch-ccl exchange
zhangxiaoli73 Jun 17, 2025
ef9ca8d
remove unneeded
zhangxiaoli73 Jun 17, 2025
997e26b
fix a bug and move to local IPC exchange
zhangxiaoli73 Jun 19, 2025
a350b8f
add symm copy_buffer API
zhangxiaoli73 Jun 19, 2025
49ca539
barrier with MPI
zhangxiaoli73 Jun 23, 2025
c0b600b
support arc
zhangxiaoli73 Jun 24, 2025
abade52
refine barrier
zhangxiaoli73 Jun 24, 2025
eb722a7
workaroud barrier
zhangxiaoli73 Jun 25, 2025
5d24272
refine ipc exchange
zhangxiaoli73 Jul 3, 2025
3be5ebe
refine ipc exchange
zhangxiaoli73 Jul 10, 2025
3d1491c
reabse and then workaround barrier
zhangxiaoli73 Jul 15, 2025
fdf3d6d
impl base class virtual function
zhangxiaoli73 Jul 17, 2025
6a7057b
format
Chao1Han Jul 31, 2025
340113f
rm hardcode ze
Chao1Han Jul 31, 2025
769ddec
clean code
Chao1Han Aug 5, 2025
4d7e170
update
Chao1Han Aug 7, 2025
de4da04
avoid symbol conflict
Chao1Han Aug 7, 2025
ffd3a90
refine IPCExchang
Chao1Han Aug 8, 2025
66b77b1
rm header
Chao1Han Aug 8, 2025
868cae8
update
Chao1Han Aug 12, 2025
f3f6f0c
refine and add sycl signal
Chao1Han Aug 18, 2025
f60645e
correct rendezvous
Chao1Han Aug 25, 2025
2b6e84f
add barrier imple by dist allgather
Chao1Han Sep 2, 2025
1df3f9f
rename ze_symbol
Chao1Han Sep 4, 2025
4034168
add allreduce_barrier
Chao1Han Sep 8, 2025
1bb3f9b
correct
Chao1Han Sep 9, 2025
34ad35f
add test case
Chao1Han Sep 9, 2025
7f0524a
rm signal
Chao1Han Sep 9, 2025
3b3ff27
code refine
zhangxiaoli73 Sep 9, 2025
0477394
refine
Chao1Han Sep 10, 2025
7788d4d
remove level-zero header if it's no needed
zhangxiaoli73 Sep 10, 2025
62bb559
fix and add test
Chao1Han Sep 11, 2025
902a61f
fix memory release issue
zhangxiaoli73 Sep 12, 2025
d301bdf
refine according to latest pytorch
zhangxiaoli73 Sep 12, 2025
c417b29
debug about getting ptr device
zhangxiaoli73 Sep 12, 2025
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions src/ATen/native/xpu/Copy.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -71,8 +71,10 @@ void memcpyAsync(
Device dst_device = iter.device(0);
Device src_device = iter.device(1);
if (dst_device == src_device) {
std::cout << "zl_debug: go to same device and specialized kernel" << std::endl;
copy_kernel(iter);
} else {
std::cout << "zl_debug: go to sycl copy kernel" << std::endl;
TORCH_INTERNAL_ASSERT(p2p_enabled == true);
auto dst = (char*)iter.data_ptr(0);
auto src = (char*)iter.data_ptr(1);
Expand Down
Loading
Loading