February 11

On Monday, 10 February 2025 at 12:17:42 UTC, FeepingCreature wrote:

>

NEW GC NEW GC NEW GC! I tried it on an internal service that's been known as a bit of a memory hog.

Not as fast as the current one right now as far as I can see. A memory heavy test workload seems about 2x slower.

But tbh I'll take that because:

Stock LDC based service after a memory hungry operation: 2952MB res
SDC/Symmetry GC after the same operation: 907MB res

That's genuinely amazing to me. Very excited for how this thing will run with a bit more opt.

Correction: after more testing this seems to be almost entirely due to use of --link-defaultlib-shared. The new GC seems about as fast as the old one given the same flags.

4 days ago

On Tuesday, 11 February 2025 at 16:30:57 UTC, FeepingCreature wrote:

>

On Monday, 10 February 2025 at 12:17:42 UTC, FeepingCreature wrote:

>

NEW GC NEW GC NEW GC! I tried it on an internal service that's been known as a bit of a memory hog.

Not as fast as the current one right now as far as I can see. A memory heavy test workload seems about 2x slower.

But tbh I'll take that because:

Stock LDC based service after a memory hungry operation: 2952MB res
SDC/Symmetry GC after the same operation: 907MB res

That's genuinely amazing to me. Very excited for how this thing will run with a bit more opt.

Correction: after more testing this seems to be almost entirely due to use of --link-defaultlib-shared. The new GC seems about as fast as the old one given the same flags.

On Tuesday, 11 February 2025 at 16:30:57 UTC, FeepingCreature wrote:

>

On Monday, 10 February 2025 at 12:17:42 UTC, FeepingCreature wrote:

>

NEW GC NEW GC NEW GC! I tried it on an internal service that's been known as a bit of a memory hog.

Not as fast as the current one right now as far as I can see. A memory heavy test workload seems about 2x slower.

But tbh I'll take that because:

Stock LDC based service after a memory hungry operation: 2952MB res
SDC/Symmetry GC after the same operation: 907MB res

That's genuinely amazing to me. Very excited for how this thing will run with a bit more opt.

Correction: after more testing this seems to be almost entirely due to use of --link-defaultlib-shared. The new GC seems about as fast as the old one given the same flags.

That's fantastic news!

Have you tried the parallel mark? You should be able to by nto passing --DRT-gcopt=parallel:0. That being said, there are a few pitfalls remainign with that one, so we recomand not using it in prod at this stage. Nevertheless, having feedback about it would be nice, and if it works for you, it works for you.

Note that the problems you can experience with that won't be memory corruption and alike, but incompatibilities with libraries that start their own threads, this kind of things.

1 2
Next ›   Last »