Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[src] Add optimization to cuda allocator #2820

Merged
merged 3 commits into from
Nov 15, 2018
Merged

[src] Add optimization to cuda allocator #2820

Changes from 1 commit
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
a545718
[src] Add optimization to cuda allocator
luitjens Nov 6, 2018
418b8e6
Respond to review feedback. Fix formatting.
ryanleary Nov 6, 2018
d9f0034
Rename new iterator to search_iter
ryanleary Nov 6, 2018
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
14 changes: 14 additions & 0 deletions src/cudamatrix/cu-allocator.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -185,6 +185,20 @@ void* CuMemoryAllocator::MallocFromSubregion(SubRegion *subregion,
// region was sufficiently large. We don't check this; if it segfaults, we'll
// debug.

int max_iters=20;
int i=0;
//search for a block that we don't have to synchronize on
auto it = iter;
while (it != subregion->free_blocks.end() && i<max_iters) {
ryanleary marked this conversation as resolved.
Show resolved Hide resolved
if (it->second->thread_id == std::this_thread::get_id() || it->second->t <= synchronize_gpu_t_) {
iter = it;
break;
} else {
it++;
i++;
}
}

MemoryBlock *block = iter->second;
// Erase 'block' from its subregion's free blocks list... the next lines are
// similar to RemoveFromFreeBlocks(), but we code it directly as we have the
Expand Down