Matthew Parkinson

I am going to look into a more consolidating calls to `madvise`, which will hopefully reduce this cost. So did it work in terms of reducing the memory usage, or...

@schrodingerZhu would you be able to run this experiment again with the latest main branch? I have done a lot of work on bringing down the footprint, most examples are...

On x86 the code would look something like: ```asm 00000000000055a0 : 55a0: 48 89 fa mov %rdi,%rdx 55a3: 48 89 fe mov %rdi,%rsi 55a6: 48 81 e6 00 00 f0...

Would you be able to post the precise commit you used on `snmalloc2`? I would have expected snmalloc2 to work better with Large Pages than `snmalloc1`. A lot of the...

Thanks, I have pushed a branch inspired by your code change: https://github.com/mjp41/snmalloc/commit/13ddd507b1912b38acf74a6a151a2130a573cd3c . This passes the test harness. What is the workload that was causing your crash? The code needs...

Although the designs between mimalloc and snmalloc were very similar I think we had a different bias in why we had per slab free lists. In snmalloc it was primarily...

@schrodingerZhu I think it would be great to have something like this. But I also share Daan's (@daanx) worry. Personally, I use the micro-benchmarks for finding problems rather than justifying...

Does the machine you run on have transparent hugh pages? You can check with ``` cat /sys/kernel/mm/transparent_hugepage/enabled ``` If it is set to always, I have found `snmalloc` can have...

Apologies, I broke CI with the branch name change, can you rebase onto the latest `main` to get the checks to run. Sorry, I haven't commented on this yet, I...

Thanks for looking into this. So here are some high-level points * I like having a completely distinct allocator for the concealed version. * I am not happy with the...