Auto merge of #126158 - Urgau:disallow-cfgs, r=petrochenkov

Disallow setting some built-in cfg via set the command-line

This PR disallow users from setting some built-in cfg via set the command-line in order to prevent incoherent state, eg. `windows` cfg active but target is Linux based.

This implements MCP rust-lang/compiler-team#610, with the caveat that we disallow cfgs no matter if they make sense or not, since I don't think it's useful to allow users to set a cfg that will be set anyway. It also complicates the implementation.

------

The `explicit_builtin_cfgs_in_flags` lint detects builtin cfgs set via the `--cfg` flag.

*(deny-by-default)*

### Example

```text
rustc --cfg unix
```

```rust,ignore (needs command line option)
fn main() {}
```

This will produce:

```text
error: unexpected `--cfg unix` flag
  |
  = note: config `unix` is only supposed to be controlled by `--target`
  = note: manually setting a built-in cfg can and does create incoherent behaviours
  = note: `#[deny(explicit_builtin_cfgs_in_flags)]` on by default
```

### Explanation

Setting builtin cfgs can and does produce incoherent behaviour, it's better to the use the appropriate `rustc` flag that controls the config. For example setting the `windows` cfg but on Linux based target.

-----

r? `@petrochenkov`
cc `@jyn514`

try-job: aarch64-apple
try-job: test-various
try-job: armhf-gnu
try-job: x86_64-msvc
try-job: x86_64-mingw
try-job: i686-msvc
try-job: i686-mingw
try-job: x86_64-gnu-llvm-17
try-job: dist-various-1

compiler/rustc_builtin_macros/src/format.rs

@@ -555,7 +555,7 @@ fn make_format_args(
             };
             let arg_name = args.explicit_args()[index].kind.ident().unwrap();
             ecx.buffered_early_lint.push(BufferedEarlyLint {
-                span: arg_name.span.into(),
+                span: Some(arg_name.span.into()),
                 node_id: rustc_ast::CRATE_NODE_ID,
                 lint_id: LintId::of(NAMED_ARGUMENTS_USED_POSITIONALLY),
                 diagnostic: BuiltinLintDiag::NamedArgumentUsedPositionally {

compiler/rustc_lint/messages.ftl

@@ -775,6 +775,10 @@ lint_undropped_manually_drops = calls to `std::mem::drop` with `std::mem::Manual
     .label = argument has type `{$arg_ty}`
     .suggestion = use `std::mem::ManuallyDrop::into_inner` to get the inner value
 
+lint_unexpected_builtin_cfg = unexpected `--cfg {$cfg}` flag
+    .controlled_by = config `{$cfg_name}` is only supposed to be controlled by `{$controlled_by}`
+    .incoherent = manually setting a built-in cfg can and does create incoherent behaviors
+
 lint_unexpected_cfg_add_build_rs_println = or consider adding `{$build_rs_println}` to the top of the `build.rs`
 lint_unexpected_cfg_add_cargo_feature = consider using a Cargo feature instead
 lint_unexpected_cfg_add_cargo_toml_lint_cfg = or consider adding in `Cargo.toml` the `check-cfg` lint config for the lint:{$cargo_toml_lint_cfg}

compiler/rustc_lint/src/context.rs

@@ -533,7 +533,7 @@ pub struct EarlyContext<'a> {
 }
 
 impl EarlyContext<'_> {
-    /// Emit a lint at the appropriate level, with an optional associated span and an existing
+    /// Emit a lint at the appropriate level, with an associated span and an existing
     /// diagnostic.
     ///
     /// [`lint_level`]: rustc_middle::lint::lint_level#decorate-signature
@@ -544,7 +544,21 @@ impl EarlyContext<'_> {
         span: MultiSpan,
         diagnostic: BuiltinLintDiag,
     ) {
-        self.opt_span_lint(lint, Some(span), |diag| {
+        self.opt_span_lint_with_diagnostics(lint, Some(span), diagnostic);
+    }
+
+    /// Emit a lint at the appropriate level, with an optional associated span and an existing
+    /// diagnostic.
+    ///
+    /// [`lint_level`]: rustc_middle::lint::lint_level#decorate-signature
+    #[rustc_lint_diagnostics]
+    pub fn opt_span_lint_with_diagnostics(
+        &self,
+        lint: &'static Lint,
+        span: Option<MultiSpan>,
+        diagnostic: BuiltinLintDiag,
+    ) {
+        self.opt_span_lint(lint, span, |diag| {
             diagnostics::decorate_lint(self.sess(), diagnostic, diag);
         });
     }

compiler/rustc_lint/src/context/diagnostics.rs

@@ -438,5 +438,8 @@ pub(super) fn decorate_lint(sess: &Session, diagnostic: BuiltinLintDiag, diag: &
         BuiltinLintDiag::OutOfScopeMacroCalls { path } => {
             lints::OutOfScopeMacroCalls { path }.decorate_lint(diag)
         }
+        BuiltinLintDiag::UnexpectedBuiltinCfg { cfg, cfg_name, controlled_by } => {
+            lints::UnexpectedBuiltinCfg { cfg, cfg_name, controlled_by }.decorate_lint(diag)
+        }
     }
 }

compiler/rustc_lint/src/early.rs

@@ -47,7 +47,7 @@ impl<'a, T: EarlyLintPass> EarlyContextAndPass<'a, T> {
     fn inlined_check_id(&mut self, id: ast::NodeId) {
         for early_lint in self.context.buffered.take(id) {
             let BufferedEarlyLint { span, node_id: _, lint_id, diagnostic } = early_lint;
-            self.context.span_lint_with_diagnostics(lint_id.lint, span, diagnostic);
+            self.context.opt_span_lint_with_diagnostics(lint_id.lint, span, diagnostic);
         }
     }
 

compiler/rustc_lint/src/lints.rs

@@ -2379,6 +2379,16 @@ pub mod unexpected_cfg_value {
     }
 }
 
+#[derive(LintDiagnostic)]
+#[diag(lint_unexpected_builtin_cfg)]
+#[note(lint_controlled_by)]
+#[note(lint_incoherent)]
+pub struct UnexpectedBuiltinCfg {
+    pub(crate) cfg: String,
+    pub(crate) cfg_name: Symbol,
+    pub(crate) controlled_by: &'static str,
+}
+
 #[derive(LintDiagnostic)]
 #[diag(lint_macro_use_deprecated)]
 #[help]

compiler/rustc_lint_defs/src/builtin.rs

@@ -42,6 +42,7 @@ declare_lint_pass! {
         DUPLICATE_MACRO_ATTRIBUTES,
         ELIDED_LIFETIMES_IN_ASSOCIATED_CONSTANT,
         ELIDED_LIFETIMES_IN_PATHS,
+        EXPLICIT_BUILTIN_CFGS_IN_FLAGS,
         EXPORTED_PRIVATE_DEPENDENCIES,
         FFI_UNWIND_CALLS,
         FORBIDDEN_LINT_GROUPS,
@@ -3288,6 +3289,39 @@ declare_lint! {
     "detects unexpected names and values in `#[cfg]` conditions",
 }
 
+declare_lint! {
+    /// The `explicit_builtin_cfgs_in_flags` lint detects builtin cfgs set via the `--cfg` flag.
+    ///
+    /// ### Example
+    ///
+    /// ```text
+    /// rustc --cfg unix
+    /// ```
+    ///
+    /// ```rust,ignore (needs command line option)
+    /// fn main() {}
+    /// ```
+    ///
+    /// This will produce:
+    ///
+    /// ```text
+    /// error: unexpected `--cfg unix` flag
+    ///   |
+    ///   = note: config `unix` is only supposed to be controlled by `--target`
+    ///   = note: manually setting a built-in cfg can and does create incoherent behaviors
+    ///   = note: `#[deny(explicit_builtin_cfgs_in_flags)]` on by default
+    /// ```
+    ///
+    /// ### Explanation
+    ///
+    /// Setting builtin cfgs can and does produce incoherent behavior, it's better to the use
+    /// the appropriate `rustc` flag that controls the config. For example setting the `windows`
+    /// cfg but on Linux based target.
+    pub EXPLICIT_BUILTIN_CFGS_IN_FLAGS,
+    Deny,
+    "detects builtin cfgs set via the `--cfg`"
+}
+
 declare_lint! {
     /// The `repr_transparent_external_private_fields` lint
     /// detects types marked `#[repr(transparent)]` that (transitively)

compiler/rustc_lint_defs/src/lib.rs

@@ -746,14 +746,19 @@ pub enum BuiltinLintDiag {
     OutOfScopeMacroCalls {
         path: String,
     },
+    UnexpectedBuiltinCfg {
+        cfg: String,
+        cfg_name: Symbol,
+        controlled_by: &'static str,
+    },
 }
 
 /// Lints that are buffered up early on in the `Session` before the
 /// `LintLevels` is calculated.
 #[derive(Debug)]
 pub struct BufferedEarlyLint {
     /// The span of code that we are linting on.
-    pub span: MultiSpan,
+    pub span: Option<MultiSpan>,
 
     /// The `NodeId` of the AST node that generated the lint.
     pub node_id: NodeId,
@@ -791,7 +796,7 @@ impl LintBuffer {
         self.add_early_lint(BufferedEarlyLint {
             lint_id: LintId::of(lint),
             node_id,
-            span: span.into(),
+            span: Some(span.into()),
             diagnostic,
         });
     }

compiler/rustc_session/src/config.rs

@@ -1304,7 +1304,10 @@ pub(crate) const fn default_lib_output() -> CrateType {
 }
 
 pub fn build_configuration(sess: &Session, mut user_cfg: Cfg) -> Cfg {
-    // Combine the configuration requested by the session (command line) with
+    // First disallow some configuration given on the command line
+    cfg::disallow_cfgs(sess, &user_cfg);
+
+    // Then combine the configuration requested by the session (command line) with
     // some default and generated configuration items.
     user_cfg.extend(cfg::default_configuration(sess));
     user_cfg

compiler/rustc_session/src/config/cfg.rs

@@ -17,12 +17,16 @@
 //!  - Add the activation logic in [`default_configuration`]
 //!  - Add the cfg to [`CheckCfg::fill_well_known`] (and related files),
 //!    so that the compiler can know the cfg is expected
+//!  - Add the cfg in [`disallow_cfgs`] to disallow users from setting it via `--cfg`
 //!  - Add the feature gating in `compiler/rustc_feature/src/builtin_attrs.rs`
 
 use std::hash::Hash;
 use std::iter;
 
+use rustc_ast::ast;
 use rustc_data_structures::fx::{FxHashMap, FxHashSet, FxIndexSet};
+use rustc_lint_defs::builtin::EXPLICIT_BUILTIN_CFGS_IN_FLAGS;
+use rustc_lint_defs::BuiltinLintDiag;
 use rustc_span::symbol::{sym, Symbol};
 use rustc_target::abi::Align;
 use rustc_target::spec::{PanicStrategy, RelocModel, SanitizerSet, Target, TargetTriple, TARGETS};
@@ -82,6 +86,67 @@ impl<'a, T: Eq + Hash + Copy + 'a> Extend<&'a T> for ExpectedValues<T> {
     }
 }
 
+/// Disallow builtin cfgs from the CLI.
+pub(crate) fn disallow_cfgs(sess: &Session, user_cfgs: &Cfg) {
+    let disallow = |cfg: &(Symbol, Option<Symbol>), controlled_by| {
+        let cfg_name = cfg.0;
+        let cfg = if let Some(value) = cfg.1 {
+            format!(r#"{}="{}""#, cfg_name, value)
+        } else {
+            format!("{}", cfg_name)
+        };
+        sess.psess.opt_span_buffer_lint(
+            EXPLICIT_BUILTIN_CFGS_IN_FLAGS,
+            None,
+            ast::CRATE_NODE_ID,
+            BuiltinLintDiag::UnexpectedBuiltinCfg { cfg, cfg_name, controlled_by },
+        )
+    };
+
+    // We want to restrict setting builtin cfgs that will produce incoherent behavior
+    // between the cfg and the rustc cli flag that sets it.
+    //
+    // The tests are in tests/ui/cfg/disallowed-cli-cfgs.rs.
+
+    // By-default all builtin cfgs are disallowed, only those are allowed:
+    //  - test: as it makes sense to the have the `test` cfg active without the builtin
+    //          test harness. See Cargo `harness = false` config.
+    //
+    // Cargo `--cfg test`: https://github.com/rust-lang/cargo/blob/bc89bffa5987d4af8f71011c7557119b39e44a65/src/cargo/core/compiler/mod.rs#L1124
+
+    for cfg in user_cfgs {
+        match cfg {
+            (sym::overflow_checks, None) => disallow(cfg, "-C overflow-checks"),
+            (sym::debug_assertions, None) => disallow(cfg, "-C debug-assertions"),
+            (sym::ub_checks, None) => disallow(cfg, "-Z ub-checks"),
+            (sym::sanitize, None | Some(_)) => disallow(cfg, "-Z sanitizer"),
+            (
+                sym::sanitizer_cfi_generalize_pointers | sym::sanitizer_cfi_normalize_integers,
+                None | Some(_),
+            ) => disallow(cfg, "-Z sanitizer=cfi"),
+            (sym::proc_macro, None) => disallow(cfg, "--crate-type proc-macro"),
+            (sym::panic, Some(sym::abort | sym::unwind)) => disallow(cfg, "-C panic"),
+            (sym::target_feature, Some(_)) => disallow(cfg, "-C target-feature"),
+            (sym::unix, None)
+            | (sym::windows, None)
+            | (sym::relocation_model, Some(_))
+            | (sym::target_abi, None | Some(_))
+            | (sym::target_arch, Some(_))
+            | (sym::target_endian, Some(_))
+            | (sym::target_env, None | Some(_))
+            | (sym::target_family, Some(_))
+            | (sym::target_os, Some(_))
+            | (sym::target_pointer_width, Some(_))
+            | (sym::target_vendor, None | Some(_))
+            | (sym::target_has_atomic, Some(_))
+            | (sym::target_has_atomic_equal_alignment, Some(_))
+            | (sym::target_has_atomic_load_store, Some(_))
+            | (sym::target_thread_local, None) => disallow(cfg, "--target"),
+            _ => {}
+        }
+    }
+}
+
 /// Generate the default configs for a given session
 pub(crate) fn default_configuration(sess: &Session) -> Cfg {
     let mut ret = Cfg::default();

compiler/rustc_session/src/parse.rs

@@ -306,10 +306,20 @@ impl ParseSess {
         span: impl Into<MultiSpan>,
         node_id: NodeId,
         diagnostic: BuiltinLintDiag,
+    ) {
+        self.opt_span_buffer_lint(lint, Some(span.into()), node_id, diagnostic)
+    }
+
+    pub fn opt_span_buffer_lint(
+        &self,
+        lint: &'static Lint,
+        span: Option<MultiSpan>,
+        node_id: NodeId,
+        diagnostic: BuiltinLintDiag,
     ) {
         self.buffered_lints.with_lock(|buffered_lints| {
             buffered_lints.push(BufferedEarlyLint {
-                span: span.into(),
+                span,
                 node_id,
                 lint_id: LintId::of(lint),
                 diagnostic,

tests/assembly/x86-return-float.rs

@@ -8,9 +8,9 @@
 // Force frame pointers to make ASM more consistent between targets
 //@ compile-flags: -O -C force-frame-pointers
 //@ filecheck-flags: --implicit-check-not fld --implicit-check-not fst
-//@ revisions: unix windows
-//@[unix] ignore-windows
-//@[windows] only-windows
+//@ revisions: normal win
+//@[normal] ignore-windows
+//@[win] only-windows
 
 #![crate_type = "lib"]
 #![feature(f16, f128)]
@@ -190,10 +190,10 @@ pub unsafe fn call_f64_f64(x: &mut (f64, f64)) {
     }
     // CHECK: movl {{.*}}(%ebp), %[[PTR:.*]]
     // CHECK: calll {{()|_}}get_f64_f64
-    // unix: movsd [[#%d,OFFSET:]](%ebp), %[[VAL1:.*]]
-    // unix-NEXT: movsd [[#%d,OFFSET+8]](%ebp), %[[VAL2:.*]]
-    // windows: movsd (%esp), %[[VAL1:.*]]
-    // windows-NEXT: movsd 8(%esp), %[[VAL2:.*]]
+    // normal: movsd [[#%d,OFFSET:]](%ebp), %[[VAL1:.*]]
+    // normal-NEXT: movsd [[#%d,OFFSET+8]](%ebp), %[[VAL2:.*]]
+    // win: movsd (%esp), %[[VAL1:.*]]
+    // win-NEXT: movsd 8(%esp), %[[VAL2:.*]]
     // CHECK-NEXT: movsd %[[VAL1]], (%[[PTR]])
     // CHECK-NEXT: movsd %[[VAL2]], 8(%[[PTR]])
     *x = get_f64_f64();
@@ -207,13 +207,13 @@ pub unsafe fn call_f32_f64(x: &mut (f32, f64)) {
     }
     // CHECK: movl {{.*}}(%ebp), %[[PTR:.*]]
     // CHECK: calll {{()|_}}get_f32_f64
-    // unix: movss [[#%d,OFFSET:]](%ebp), %[[VAL1:.*]]
-    // unix-NEXT: movsd [[#%d,OFFSET+4]](%ebp), %[[VAL2:.*]]
-    // windows: movss (%esp), %[[VAL1:.*]]
-    // windows-NEXT: movsd 8(%esp), %[[VAL2:.*]]
+    // normal: movss [[#%d,OFFSET:]](%ebp), %[[VAL1:.*]]
+    // normal-NEXT: movsd [[#%d,OFFSET+4]](%ebp), %[[VAL2:.*]]
+    // win: movss (%esp), %[[VAL1:.*]]
+    // win-NEXT: movsd 8(%esp), %[[VAL2:.*]]
     // CHECK-NEXT: movss %[[VAL1]], (%[[PTR]])
-    // unix-NEXT: movsd %[[VAL2]], 4(%[[PTR]])
-    // windows-NEXT: movsd %[[VAL2]], 8(%[[PTR]])
+    // normal-NEXT: movsd %[[VAL2]], 4(%[[PTR]])
+    // win-NEXT: movsd %[[VAL2]], 8(%[[PTR]])
     *x = get_f32_f64();
 }
 
@@ -225,10 +225,10 @@ pub unsafe fn call_f64_f32(x: &mut (f64, f32)) {
     }
     // CHECK: movl {{.*}}(%ebp), %[[PTR:.*]]
     // CHECK: calll {{()|_}}get_f64_f32
-    // unix: movsd [[#%d,OFFSET:]](%ebp), %[[VAL1:.*]]
-    // unix-NEXT: movss [[#%d,OFFSET+8]](%ebp), %[[VAL2:.*]]
-    // windows: movsd (%esp), %[[VAL1:.*]]
-    // windows-NEXT: movss 8(%esp), %[[VAL2:.*]]
+    // normal: movsd [[#%d,OFFSET:]](%ebp), %[[VAL1:.*]]
+    // normal-NEXT: movss [[#%d,OFFSET+8]](%ebp), %[[VAL2:.*]]
+    // win: movsd (%esp), %[[VAL1:.*]]
+    // win-NEXT: movss 8(%esp), %[[VAL2:.*]]
     // CHECK-NEXT: movsd %[[VAL1]], (%[[PTR]])
     // CHECK-NEXT: movss %[[VAL2]], 8(%[[PTR]])
     *x = get_f64_f32();
@@ -257,10 +257,10 @@ pub unsafe fn call_f64_other(x: &mut (f64, usize)) {
     }
     // CHECK: movl {{.*}}(%ebp), %[[PTR:.*]]
     // CHECK: calll {{()|_}}get_f64_other
-    // unix: movsd [[#%d,OFFSET:]](%ebp), %[[VAL1:.*]]
-    // unix-NEXT: movl [[#%d,OFFSET+8]](%ebp), %[[VAL2:.*]]
-    // windows: movsd (%esp), %[[VAL1:.*]]
-    // windows-NEXT: movl 8(%esp), %[[VAL2:.*]]
+    // normal: movsd [[#%d,OFFSET:]](%ebp), %[[VAL1:.*]]
+    // normal-NEXT: movl [[#%d,OFFSET+8]](%ebp), %[[VAL2:.*]]
+    // win: movsd (%esp), %[[VAL1:.*]]
+    // win-NEXT: movl 8(%esp), %[[VAL2:.*]]
     // CHECK-NEXT: movsd %[[VAL1]], (%[[PTR]])
     // CHECK-NEXT: movl %[[VAL2]], 8(%[[PTR]])
     *x = get_f64_other();
@@ -289,13 +289,13 @@ pub unsafe fn call_other_f64(x: &mut (usize, f64)) {
     }
     // CHECK: movl {{.*}}(%ebp), %[[PTR:.*]]
     // CHECK: calll {{()|_}}get_other_f64
-    // unix: movl [[#%d,OFFSET:]](%ebp), %[[VAL1:.*]]
-    // unix-NEXT: movsd [[#%d,OFFSET+4]](%ebp), %[[VAL2:.*]]
-    // windows: movl (%esp), %[[VAL1:.*]]
-    // windows-NEXT: movsd 8(%esp), %[[VAL2:.*]]
+    // normal: movl [[#%d,OFFSET:]](%ebp), %[[VAL1:.*]]
+    // normal-NEXT: movsd [[#%d,OFFSET+4]](%ebp), %[[VAL2:.*]]
+    // win: movl (%esp), %[[VAL1:.*]]
+    // win-NEXT: movsd 8(%esp), %[[VAL2:.*]]
     // CHECK-NEXT: movl %[[VAL1]], (%[[PTR]])
-    // unix-NEXT: movsd %[[VAL2]], 4(%[[PTR]])
-    // windows-NEXT: movsd %[[VAL2]], 8(%[[PTR]])
+    // normal-NEXT: movsd %[[VAL2]], 4(%[[PTR]])
+    // win-NEXT: movsd %[[VAL2]], 8(%[[PTR]])
     *x = get_other_f64();
 }