Skip to content

Document variance in strip #13677

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Document variance in strip #13677

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 4 additions & 2 deletions src/doc/src/reference/profiles.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -117,7 +117,9 @@ strip = "debuginfo"
```

Possible string values of `strip` are `"none"`, `"debuginfo"`, and `"symbols"`.
The default is `"none"`.
The default is `"none"` by default, and `"debuginfo"` in release profile.
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Similar to #13638, I don't think this is entirely correct. The default strip behavior depends on if debug is set (including overrides).

In #13257, we somewhat intentionally did not document this, as it should be a completely internal matter that should be transparent to the user. It is only running strip to deal with the particular way the standard library is distributed.

I'm not yet entirely convinced this should be documented, since it is a relatively complicated thing to explain to the user (something they shouldn't even need to know about) which could just contribute to confusion.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

...Frankly, I don't know why there are even two options if one affects how the other is interpreted.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

In any case, I do not believe these behaviors are actually completely internal matters that are transparent to the user, as the effective behavior seems to change wildly based on the target, host, or available binaries, e.g.

Perhaps we can make it one with rust-lang/rust#123151 but it isn't now.

Copy link
Contributor

@orlp orlp Jan 2, 2025 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@ehuss The current default lying about being "none" made it much harder today to debug a problem where the compiler is stripping all symbols on MacOS if debug = false is specified: rust-lang/rust#135028.

I agree with you however that the wording by @workingjubilee is not entirely correct, the default does not depend on the profile, it depends on whether debug is specified.

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@ehuss What should the behavior be when we replace a strip binary with Rust library code?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

There isn't actually anything "transparent" here when they can set these options, examine a backtrace, and then set these options again, and examine a different backtrace.

Be aware that stripping `"symbols"` may make backtraces incomprehensible,
depending on the platform.
Comment on lines +121 to +122
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

For the most part, the cargo documentation doesn't dive into the particulars of explaining the different options and their consequences. That documentation should generally live in the rustc book (so we don't have to duplicate it in two remote places). Adding warnings about the consequences of stripping symbols would probably be appropriate over there (and seems good to include).

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It is my experience that people often do not actually click through, simply put.


You can also configure this option with the boolean values `true` or `false`.
`strip = true` is equivalent to `strip = "symbols"`. `strip = false` is
Expand Down Expand Up @@ -293,7 +295,7 @@ The default settings for the `release` profile are:
opt-level = 3
debug = false
split-debuginfo = '...' # Platform-specific.
strip = "none"
strip = "debuginfo"
debug-assertions = false
overflow-checks = false
lto = false
Expand Down