From ced8d311a593fcf9c3823e4e118474ac132d8e60 Mon Sep 17 00:00:00 2001
From: Picnoir <picnoir@alternativebit.fr>
Date: Wed, 20 Nov 2024 13:33:00 +0100
Subject: [PATCH] gc: resume GC after a pathinuse error

First the motivation: I recently faced a bug that I assume is coming
from the topoSortPaths function where the GC was trying to delete a
path having some alive referrers. I resolved this by manually deleting
the faulty path referrers using nix-store --query --referrers. I sadly
did not manage to reproduce this bug.

This bug alone is not a big deal. However, this bug is
triggering a cascading failure: invalidatePathChecked is throwing a
PathInUse exception. This exception is not catched and fails the whole GC
run. From there, the machine (a builder machine) was unable to GC its
Nix store, which led to an almost full disk with no way to
automatically delete the dead Nix paths.

Instead, I think we should log the error for the specific store path
we're trying to delete, specifying we can't delete this path because
it still has referrers. Once we're done with logging that, the GC run
should continue to delete the dead store paths it can delete.
---
 src/libstore/gc.cc | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)

diff --git a/src/libstore/gc.cc b/src/libstore/gc.cc
index 73195794a..45dfe4ad8 100644
--- a/src/libstore/gc.cc
+++ b/src/libstore/gc.cc
@@ -4,6 +4,7 @@
 #include "finally.hh"
 #include "unix-domain-socket.hh"
 #include "signals.hh"
+#include "posix-fs-canonicalise.hh"
 
 #if !defined(__linux__)
 // For shelling out to lsof
@@ -763,13 +764,18 @@ void LocalStore::collectGarbage(const GCOptions & options, GCResults & results)
                 }
             }
         }
-
         for (auto & path : topoSortPaths(visited)) {
             if (!dead.insert(path).second) continue;
             if (shouldDelete) {
-                invalidatePathChecked(path);
-                deleteFromStore(path.to_string());
-                referrersCache.erase(path);
+                try {
+                    invalidatePathChecked(path);
+                    deleteFromStore(path.to_string());
+                    referrersCache.erase(path);
+                } catch (PathInUse &e) {
+                    // If we end up here, it's likely a new occurence
+                    // of https://github.com/NixOS/nix/issues/11923
+                    printError("BUG: %s", e.what());
+                }
             }
         }
     };