tarfs reliability fixes #160
Merged
+33
−24
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
If we don't keep it up to date, d_off will be incorrect, so callers will seek to the wrong position if they ever try to use d_off. Signed-off-by: Wedson Almeida Filho <walmeida@microsoft.com>
If we run into an error while listing the contents of a directory, don't fail the whole request if we already emitted at least one entry. Signed-off-by: Wedson Almeida Filho <walmeida@microsoft.com>
Without this, the lock would stay locked indefinitely. Signed-off-by: Wedson Almeida Filho <walmeida@microsoft.com>
These checks will never trigger for valid images but they are useful when dealing with corrupted or malicious images: although not strictly necessary as block devices should reject such read attempts, having them is a good security-in-depth approach because it protects the block layer. Signed-off-by: Wedson Almeida Filho <walmeida@microsoft.com>
This the right function to call on failure because it unhashes the inode, which will cause any waiters (e.g., callers of ilookup or iget_locked) to retry after waiting for inode init to complete. Signed-off-by: Wedson Almeida Filho <walmeida@microsoft.com>
This check ensures that the expression below to determine the superblock position won't underflow. Signed-off-by: Wedson Almeida Filho <walmeida@microsoft.com>
When a mounted filesystem needs memory that isn't available, this flag instructs the kernel allocator to not request (and wait for) file systems to shrink caches because it may lead to deadlocks. Signed-off-by: Wedson Almeida Filho <walmeida@microsoft.com>
wedsonaf
added
the
upstream/not-needed
dallasd1
approved these changes
Feb 12, 2024
miz060
approved these changes
Feb 12, 2024
|
Should this patch set be added here or is that just a historical archive? |
Hey Malte, That’s our archive upstream tarfs repo which hasn’t been updated recently. @wedsonaf is actively involved in the process of upstreaming the tarfs module to the main Linux kernel. For the most recent updates and fixes to tarfs, we always push them to the msft-main branch for internal testing first. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Merge Checklist
upstream-missinglabel (orupstream-not-needed) has been set on the PR.Summary
Fixes related to rarely exercised paths like failure or partial failure paths, tiny block devices (e.g., size of a single sector), corrupted images, etc.
Upstreaming to kata-containers is not necessary because we are hosting this in a different repo.
Test Methodology
Built the kernel module locally in 5.15 and 6.1 kernels, booted an image with busybox, loaded the module, mounted a tarfs file system and used it.