Avoid exponential explosion with pre-resolved inline frames
- Dominant language
- Python
- Stars
- 44.8k
- Forks
- 4.9k
- Avg merge
- 1d 5h
- Merged PRs (30d)
- 624
Description
### Environment
SaaS (https://sentry.io/)
### Version
_No response_
### Link
_No response_
### DSN
_No response_
### Steps to Reproduce
I wrote a detailed analysis on https://github.com/getsentry/sentry-rust/issues/470#issuecomment-1472136215.
The problem is that a native executable that does full symbolication with expansion of inline functions on the client/SDK side, and sends those expanded inline frames to the server.
For each of the already-expanded inline frames that share the same instruction addr per definition, will be expanded *again* when symbolicating server-side, leading to exponential grows in the number of output frames, and it is super confusing to customers as well.
### Expected Result
One "physical" frame should expand to N inline frames.
### Actual Result
One "physical" frame is expanded to N inline frames on the SDK, which are then again expanded to N*N inline frames after server-side symbolication.
Contributor guide
Assessment
This issue has not been assessed yet.