feat(tvix/store): support RUST_LOG env var r/7723

This allows selectively increasing the log level for only parts of the
stack.

For example, the following RUST_LOG env var enables "tracing" level
logging for `tvix_store` and `tvix_castore`, while keeping it at "info"
for the rest of the stack:

export RUST_LOG='info,tvix_store=trace,tvix_castore=trace'

It only affects logs, not traces (if enabled).

Change-Id: Ib936bd132a405f216e75c843db83fbd71d20a18a
Reviewed-on: https://cl.tvl.fyi/c/depot/+/11182
Autosubmit: flokli <flokli@flokli.de>
Tested-by: BuildkiteCI
Reviewed-by: Connor Brewster <cbrewster@hey.com>

4 files changed, 153 insertions, 20 deletions

diff --git a/tvix/Cargo.lock b/tvix/Cargo.lock
index a10dabdc0ee0..ec5955188421 100644
--- a/tvix/Cargo.lock
+++ b/tvix/Cargo.lock
@@ -385,7 +385,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "c48f0051a4b4c5e0b6d365cd04af53aeaa209e3cc15ec2cdb69e73cc87fbd0dc"
 dependencies = [
  "memchr",
- "regex-automata",
+ "regex-automata 0.4.3",
  "serde",
 ]
 
@@ -1711,6 +1711,15 @@ dependencies = [
 ]
 
 [[package]]
+name = "matchers"
+version = "0.1.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "8263075bb86c5a1b1427b5ae862e8889656f126e9f77c484496e8b47cf5c5558"
+dependencies = [
+ "regex-automata 0.1.10",
+]
+
+[[package]]
 name = "matchit"
 version = "0.7.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -2291,7 +2300,7 @@ dependencies = [
  "rand",
  "rand_chacha",
  "rand_xorshift",
- "regex-syntax",
+ "regex-syntax 0.8.2",
  "rusty-fork",
  "tempfile",
  "unarray",
@@ -2532,8 +2541,17 @@ checksum = "380b951a9c5e80ddfd6136919eef32310721aa4aacd4889a8d39124b026ab343"
 dependencies = [
  "aho-corasick",
  "memchr",
- "regex-automata",
- "regex-syntax",
+ "regex-automata 0.4.3",
+ "regex-syntax 0.8.2",
+]
+
+[[package]]
+name = "regex-automata"
+version = "0.1.10"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "6c230d73fb8d8c1b9c0b3135c5142a8acee3a0558fb8db5cf1cb65f8d7862132"
+dependencies = [
+ "regex-syntax 0.6.29",
 ]
 
 [[package]]
@@ -2544,11 +2562,17 @@ checksum = "5f804c7828047e88b2d32e2d7fe5a105da8ee3264f01902f796c8e067dc2483f"
 dependencies = [
  "aho-corasick",
  "memchr",
- "regex-syntax",
+ "regex-syntax 0.8.2",
 ]
 
 [[package]]
 name = "regex-syntax"
+version = "0.6.29"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "f162c6dd7b008981e4d40210aca20b4bd0f9b60ca9271061b07f78537722f2e1"
+
+[[package]]
+name = "regex-syntax"
 version = "0.8.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "c08c74e62047bb2de4ff487b251e4a92e24f48745648451635cec7d591162d9f"
@@ -3734,12 +3758,16 @@ version = "0.3.18"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "ad0f048c97dbd9faa9b7df56362b8ebcaa52adb06b498c050d2f4e32f90a7a8b"
 dependencies = [
+ "matchers",
  "nu-ansi-term",
+ "once_cell",
+ "regex",
  "serde",
  "serde_json",
  "sharded-slab",
  "smallvec",
  "thread_local",
+ "tracing",
  "tracing-core",
  "tracing-log",
  "tracing-serde",
diff --git a/tvix/Cargo.nix b/tvix/Cargo.nix
index 87f9bbac7090..31d6faec252d 100644
--- a/tvix/Cargo.nix
+++ b/tvix/Cargo.nix
@@ -1333,7 +1333,7 @@ rec {
           }
           {
             name = "regex-automata";
-            packageId = "regex-automata";
+            packageId = "regex-automata 0.4.3";
             optional = true;
             usesDefaultFeatures = false;
             features = [ "dfa-search" ];
@@ -5264,6 +5264,22 @@ rec {
         };
         resolvedDefaultFeatures = [ "default" "v5-04" "v5-05" "v5-10" "v5-13" "v5-20" "v5-21" "v5-22" "v5-23" "v5-25" "v5-27" "v5-32" "v5-35" "v5-38" ];
       };
+      "matchers" = rec {
+        crateName = "matchers";
+        version = "0.1.0";
+        edition = "2018";
+        sha256 = "0n2mbk7lg2vf962c8xwzdq96yrc9i0p8dbmm4wa1nnkcp1dhfqw2";
+        authors = [
+          "Eliza Weisman <eliza@buoyant.io>"
+        ];
+        dependencies = [
+          {
+            name = "regex-automata";
+            packageId = "regex-automata 0.1.10";
+          }
+        ];
+
+      };
       "matchit" = rec {
         crateName = "matchit";
         version = "0.7.3";
@@ -7094,7 +7110,7 @@ rec {
           }
           {
             name = "regex-syntax";
-            packageId = "regex-syntax";
+            packageId = "regex-syntax 0.8.2";
             optional = true;
           }
           {
@@ -7777,13 +7793,13 @@ rec {
           }
           {
             name = "regex-automata";
-            packageId = "regex-automata";
+            packageId = "regex-automata 0.4.3";
             usesDefaultFeatures = false;
             features = [ "alloc" "syntax" "meta" "nfa-pikevm" ];
           }
           {
             name = "regex-syntax";
-            packageId = "regex-syntax";
+            packageId = "regex-syntax 0.8.2";
             usesDefaultFeatures = false;
           }
         ];
@@ -7811,7 +7827,31 @@ rec {
         };
         resolvedDefaultFeatures = [ "default" "perf" "perf-backtrack" "perf-cache" "perf-dfa" "perf-inline" "perf-literal" "perf-onepass" "std" "unicode" "unicode-age" "unicode-bool" "unicode-case" "unicode-gencat" "unicode-perl" "unicode-script" "unicode-segment" ];
       };
-      "regex-automata" = rec {
+      "regex-automata 0.1.10" = rec {
+        crateName = "regex-automata";
+        version = "0.1.10";
+        edition = "2015";
+        sha256 = "0ci1hvbzhrfby5fdpf4ganhf7kla58acad9i1ff1p34dzdrhs8vc";
+        authors = [
+          "Andrew Gallant <jamslam@gmail.com>"
+        ];
+        dependencies = [
+          {
+            name = "regex-syntax";
+            packageId = "regex-syntax 0.6.29";
+            optional = true;
+          }
+        ];
+        features = {
+          "default" = [ "std" ];
+          "fst" = [ "dep:fst" ];
+          "regex-syntax" = [ "dep:regex-syntax" ];
+          "std" = [ "regex-syntax" ];
+          "transducer" = [ "std" "fst" ];
+        };
+        resolvedDefaultFeatures = [ "default" "regex-syntax" "std" ];
+      };
+      "regex-automata 0.4.3" = rec {
         crateName = "regex-automata";
         version = "0.4.3";
         edition = "2021";
@@ -7835,7 +7875,7 @@ rec {
           }
           {
             name = "regex-syntax";
-            packageId = "regex-syntax";
+            packageId = "regex-syntax 0.8.2";
             optional = true;
             usesDefaultFeatures = false;
           }
@@ -7871,7 +7911,21 @@ rec {
         };
         resolvedDefaultFeatures = [ "alloc" "dfa-onepass" "dfa-search" "hybrid" "meta" "nfa-backtrack" "nfa-pikevm" "nfa-thompson" "perf-inline" "perf-literal" "perf-literal-multisubstring" "perf-literal-substring" "std" "syntax" "unicode" "unicode-age" "unicode-bool" "unicode-case" "unicode-gencat" "unicode-perl" "unicode-script" "unicode-segment" "unicode-word-boundary" ];
       };
-      "regex-syntax" = rec {
+      "regex-syntax 0.6.29" = rec {
+        crateName = "regex-syntax";
+        version = "0.6.29";
+        edition = "2018";
+        sha256 = "1qgj49vm6y3zn1hi09x91jvgkl2b1fiaq402skj83280ggfwcqpi";
+        authors = [
+          "The Rust Project Developers"
+        ];
+        features = {
+          "default" = [ "unicode" ];
+          "unicode" = [ "unicode-age" "unicode-bool" "unicode-case" "unicode-gencat" "unicode-perl" "unicode-script" "unicode-segment" ];
+        };
+        resolvedDefaultFeatures = [ "default" "unicode" "unicode-age" "unicode-bool" "unicode-case" "unicode-gencat" "unicode-perl" "unicode-script" "unicode-segment" ];
+      };
+      "regex-syntax 0.8.2" = rec {
         crateName = "regex-syntax";
         version = "0.8.2";
         edition = "2021";
@@ -11774,11 +11828,28 @@ rec {
         ];
         dependencies = [
           {
+            name = "matchers";
+            packageId = "matchers";
+            optional = true;
+          }
+          {
             name = "nu-ansi-term";
             packageId = "nu-ansi-term";
             optional = true;
           }
           {
+            name = "once_cell";
+            packageId = "once_cell";
+            optional = true;
+          }
+          {
+            name = "regex";
+            packageId = "regex";
+            optional = true;
+            usesDefaultFeatures = false;
+            features = [ "std" "unicode-case" "unicode-perl" ];
+          }
+          {
             name = "serde";
             packageId = "serde";
             optional = true;
@@ -11804,6 +11875,12 @@ rec {
             optional = true;
           }
           {
+            name = "tracing";
+            packageId = "tracing";
+            optional = true;
+            usesDefaultFeatures = false;
+          }
+          {
             name = "tracing-core";
             packageId = "tracing-core";
             usesDefaultFeatures = false;
@@ -11823,6 +11900,16 @@ rec {
         ];
         devDependencies = [
           {
+            name = "regex";
+            packageId = "regex";
+            usesDefaultFeatures = false;
+            features = [ "std" ];
+          }
+          {
+            name = "tracing";
+            packageId = "tracing";
+          }
+          {
             name = "tracing-log";
             packageId = "tracing-log";
           }
@@ -11855,7 +11942,7 @@ rec {
           "valuable-serde" = [ "dep:valuable-serde" ];
           "valuable_crate" = [ "dep:valuable_crate" ];
         };
-        resolvedDefaultFeatures = [ "alloc" "ansi" "default" "fmt" "json" "nu-ansi-term" "registry" "serde" "serde_json" "sharded-slab" "smallvec" "std" "thread_local" "tracing-log" "tracing-serde" ];
+        resolvedDefaultFeatures = [ "alloc" "ansi" "default" "env-filter" "fmt" "json" "matchers" "nu-ansi-term" "once_cell" "regex" "registry" "serde" "serde_json" "sharded-slab" "smallvec" "std" "thread_local" "tracing" "tracing-log" "tracing-serde" ];
       };
       "try-lock" = rec {
         crateName = "try-lock";
@@ -12796,7 +12883,7 @@ rec {
           {
             name = "tracing-subscriber";
             packageId = "tracing-subscriber";
-            features = [ "json" ];
+            features = [ "env-filter" "json" ];
           }
           {
             name = "tvix-castore";
diff --git a/tvix/store/Cargo.toml b/tvix/store/Cargo.toml
index 2dd155dbbc2e..366bd3921684 100644
--- a/tvix/store/Cargo.toml
+++ b/tvix/store/Cargo.toml
@@ -31,7 +31,7 @@ tonic = { version = "0.11.0", features = ["tls", "tls-roots"] }
 tower = "0.4.13"
 tracing = "0.1.37"
 tracing-opentelemetry = "0.22.0"
-tracing-subscriber = { version = "0.3.16", features = ["json"] }
+tracing-subscriber = { version = "0.3.16", features = ["env-filter", "json"] }
 tvix-castore = { path = "../castore" }
 url = "2.4.0"
 walkdir = "2.4.0"
diff --git a/tvix/store/src/bin/tvix-store.rs b/tvix/store/src/bin/tvix-store.rs
index f3a5d01b605b..27a67b7c91c1 100644
--- a/tvix/store/src/bin/tvix-store.rs
+++ b/tvix/store/src/bin/tvix-store.rs
@@ -10,7 +10,8 @@ use tokio_listener::UserOptions;
 use tonic::transport::Server;
 use tracing::info;
 use tracing::Level;
-use tracing_subscriber::fmt::writer::MakeWriterExt;
+use tracing_subscriber::EnvFilter;
+use tracing_subscriber::Layer;
 use tracing_subscriber::{layer::SubscriberExt, util::SubscriberInitExt};
 
 use tvix_castore::proto::blob_service_server::BlobServiceServer;
@@ -55,6 +56,10 @@ struct Cli {
     #[arg(long, default_missing_value = "true", default_value = "true", num_args(0..=1), require_equals(true), action(clap::ArgAction::Set))]
     otlp: bool,
 
+    /// A global log level to use when printing logs.
+    /// It's also possible to set `RUST_LOG` according to
+    /// `tracing_subscriber::filter::EnvFilter`, which will always have
+    /// priority.
     #[arg(long)]
     log_level: Option<Level>,
 
@@ -172,19 +177,32 @@ async fn main() -> Result<(), Box<dyn std::error::Error>> {
     // configure log settings
     let level = cli.log_level.unwrap_or(Level::INFO);
 
+    // Set up the tracing subscriber.
     let subscriber = tracing_subscriber::registry()
         .with(
             cli.json.then_some(
                 tracing_subscriber::fmt::Layer::new()
-                    .with_writer(std::io::stderr.with_max_level(level))
-                    .json(),
+                    .with_writer(std::io::stderr)
+                    .json()
+                    .with_filter(
+                        EnvFilter::builder()
+                            .with_default_directive(level.into())
+                            .from_env()
+                            .expect("invalid RUST_LOG"),
+                    ),
             ),
         )
         .with(
             (!cli.json).then_some(
                 tracing_subscriber::fmt::Layer::new()
-                    .with_writer(std::io::stderr.with_max_level(level))
-                    .pretty(),
+                    .with_writer(std::io::stderr)
+                    .pretty()
+                    .with_filter(
+                        EnvFilter::builder()
+                            .with_default_directive(level.into())
+                            .from_env()
+                            .expect("invalid RUST_LOG"),
+                    ),
             ),
         );