mirror of
https://github.com/lovell/sharp.git
synced 2026-01-25 16:26:05 +00:00
584807b4f5
When using the V8 memory cage, Buffers cannot be wrapped and then later freed via a callback. When the cage is detected via a throw, instead fall back to copying Buffer contents to V8 memory. This approach will be used by Electron 21+ and you should expect reduced performance and increased memory consumption/fragmentation.