Stack overflow in PrimeCaches when using PGO

rust-analyzer version: rust-analyzer version: 0.3.2433-standalone (66e3b5819e 2025-04-21) [/Users/user/.vscode/extensions/rust-lang.rust-analyzer-0.3.2433-darwin-arm64/server/rust-analyzer]

rustc version: rustc 1.88.0-nightly (78f2104e3 2025-04-16)

editor or extension: VSCode, extension 0.3.2433

Using large code base with a lot of crates in Cargo's workspace. Constantly got this in VSCode after upgrading from 0.3.2379 to 0.3.2433 (reverting back fixes this):

thread 'PrimeCaches#3' has overflowed its stack
fatal runtime error: stack overflow
[Error - 10:36:10 PM] Server process exited with signal SIGABRT.
[Info  - 10:36:11 PM] Connection to server got closed. Server will restart.
true

thread 'PrimeCaches#2' has overflowed its stack
fatal runtime error: stack overflow
[Info  - 10:36:17 PM] Connection to server got closed. Server will restart.
true
[Error - 10:36:17 PM] Server process exited with signal SIGABRT.

thread 'PrimeCaches#9' has overflowed its stack
fatal runtime error: stack overflow
[Error - 10:36:24 PM] Server process exited with signal SIGABRT.
[Info  - 10:36:24 PM] Connection to server got closed. Server will restart.
true

and so on.

Start by reproducing the stack overflow in the PrimeCaches threads with rust-analyzer 0.3.2433 and PGO, then compare the behavior with 0.3.2379, which reportedly does not fail. Investigate the change between those versions and verify that a large Cargo workspace no longer causes the server process to overflow its stack or repeatedly restart.