fix: use the gin PageAllocator instead of V8::PageAllocator #26331
Conversation
This makes browser-process JS allocate pages using the base/gin allocator thus ensuring flags such as MAP_JIT are appropriately applied.
|
Is there path to upstream these patches ? LGTM otherwise, nice find! |
The gin one, probably not. The node one definitely |
| Date: Tue, 3 Nov 2020 16:49:32 -0800 | ||
| Subject: export gin::V8Platform::PageAllocator for usage outside of the gin | ||
| platform | ||
|
|
There was a problem hiding this comment.
Can you add an explanation here w/ upstream plans?
There was a problem hiding this comment.
lint failed here :( this needs a description @MarshallOfSound
There was a problem hiding this comment.
Oops, missed this comment. How did it go green if lint was failing :/
There was a problem hiding this comment.
sorry, i meant lint failed to detect the missing description because of the two-line subject
|
Release Notes Persisted
|
Electron v12 fixes an issue in the way memory was allocated which allows us to remove an unsafe macOS codesign entitlement allowing to execute unsigned memory. See electron/electron#26331 for more details.
Electron v12 fixes an issue in the way memory was allocated which allows us to remove an unsafe macOS codesign entitlement allowing to execute unsigned memory. See electron/electron#26331 for more details.
This makes browser-process JS allocate pages using the base/gin allocator thus ensuring flags such as MAP_JIT are appropriately applied.
Without this changes apps were forced to codesign the browser process with
com.apple.security.cs.allow-unsigned-executable-memory(which is a bad thing).Notes: Updated internal memory allocation logic such that you no longer need to use the
com.apple.security.cs.allow-unsigned-executable-memorycodesign entitlement on macOS