diff options
| author | наб <nabijaczleweli@nabijaczleweli.xyz> | 2023-06-15 15:04:32 +0200 |
|---|---|---|
| committer | Luca Boccassi <luca.boccassi@gmail.com> | 2023-06-16 02:20:52 +0200 |
| commit | f26c794308975a875a98faf22a6650d7134f17c1 (patch) | |
| tree | 88bfb5da65d99a43edce77ca0b148cb019ba4f3c | |
| parent | core: fix post-fork memory leak (diff) | |
| download | systemd-f26c794308975a875a98faf22a6650d7134f17c1.tar.xz systemd-f26c794308975a875a98faf22a6650d7134f17c1.zip | |
find-esp: fix XBOOTLDR stx_dev_major=0 and not btrfs fix
The original commit is fully correct in its analysis, description, and
mechanics, but the patch changes an identical condition around line 500
(find_esp_and_warn()), instead of line 800 (find_xbootldr_and_warn()).
The internal patch I distributed to testers was correct (L800), and
neither the reviewers, nor me, caught that I wrote the wrong line for
upstream submission. I've re-checked that this patch applied to
systemd 252.11-1 fixes the issue.
Keep the -ENOTTY condition for find_esp_and_warn(), since the conditions
are clearly supposed to be the same and similar semantics apply.
Fixes: commit ed89819f8fd7bfe99cd652082076e85e1417e4e9 ("find-esp: don't
silently error bootctl install if presumed XBOOTLDR part is
stx_dev_major=0 but not btrfs")
| -rw-r--r-- | src/shared/find-esp.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/src/shared/find-esp.c b/src/shared/find-esp.c index d9336f4431..6b460a27cb 100644 --- a/src/shared/find-esp.c +++ b/src/shared/find-esp.c @@ -840,7 +840,7 @@ int find_xbootldr_and_warn_at( r = verify_xbootldr(rfd, "/boot", flags | VERIFY_ESP_SEARCHING, ret_path, ret_uuid, ret_devid); if (r < 0) { - if (!IN_SET(r, -ENOENT, -EADDRNOTAVAIL, -ENOTDIR)) /* This one is not it */ + if (!IN_SET(r, -ENOENT, -EADDRNOTAVAIL, -ENOTDIR, -ENOTTY)) /* This one is not it */ return r; return -ENOKEY; |