@@ -25,7 +25,7 @@ This will give our donors more choices for supporting Redox development.
...
@@ -25,7 +25,7 @@ This will give our donors more choices for supporting Redox development.
## Performance Improvements
## Performance Improvements
4lDO2 replaced the Redox memory manager buddy allocator with a linked-list-based O(1) frame allocator that reduced the global memory allocation overhead from 35% to 0.9%, while the overhead reduction at boot was 95%!
4lDO2 replaced the Redox memory manager buddy allocator with a linked-list-based O(1) frame allocator that reduced the global memory allocation overhead from 35% to 0.9%, while the overhead reduction at boot was 95%!
Read all about it in his excellent post "[Significant performance and correctness improvements to the kernel](/news/kernel-10.md)".
Read all about it in his excellent post "[Significant performance and correctness improvements to the kernel](/news/kernel-10)".
He also reduced the base system call overhead a lot, from 344 to 116 cycles on x86_64!
He also reduced the base system call overhead a lot, from 344 to 116 cycles on x86_64!
...
@@ -107,4 +107,4 @@ We are happy to see some progress on Redox as a web server,
...
@@ -107,4 +107,4 @@ We are happy to see some progress on Redox as a web server,
and we are very excited with the recent performance and stability improvements.
and we are very excited with the recent performance and stability improvements.
If you would like to help out, join us on [Matrix](https://matrix.to/#/#redox-join:matrix.org).
If you would like to help out, join us on [Matrix](https://matrix.to/#/#redox-join:matrix.org).