[PR #1713] [MERGED] Prevents the etag string buffer from being destroyed during use #2134

Closed
opened 2026-03-04 02:03:53 +03:00 by kerem · 0 comments
Owner

📋 Pull Request Information

Original PR: https://github.com/s3fs-fuse/s3fs-fuse/pull/1713
Author: @ggtakec
Created: 7/3/2021
Status: Merged
Merged: 7/3/2021
Merged by: @gaul

Base: masterHead: etag_stock


📝 Commits (1)

  • 0b2b7fb Prevents the etag string buffer from being destroyed during use

📊 Changes

3 files changed (+20 additions, -24 deletions)

View changed files

📝 src/fdcache_fdinfo.cpp (+12 -5)
📝 src/fdcache_fdinfo.h (+2 -18)
📝 src/types.h (+6 -1)

📄 Description

Relevant Issue (if applicable)

n/a

Details

There was a potential bug in the etag string saving process in parallel multipart uploads.
Currently, it has not been realized, but it may be a problem in future function addition(planned now), so it has been fixed.

This bug is the part that saves etag string when CURL processing is completed in parallel multipart upload.
The object has an etag as a pointer, but it is possible that the entity does not exist in the future.
Therefore, I added an etag_entities member to the PseudoFdInfo class and modified it so that the etag entity always exists for the period when the PseudoFdInfo object exists.

And the MPPART_INFO structure has almost the same purpose and member variables as the filepart structure, so it is unified as filepart.

The MPPART_INFO changes are useless after this integration, so the above fix and this integration are combined into a single(this) PR.


🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.

## 📋 Pull Request Information **Original PR:** https://github.com/s3fs-fuse/s3fs-fuse/pull/1713 **Author:** [@ggtakec](https://github.com/ggtakec) **Created:** 7/3/2021 **Status:** ✅ Merged **Merged:** 7/3/2021 **Merged by:** [@gaul](https://github.com/gaul) **Base:** `master` ← **Head:** `etag_stock` --- ### 📝 Commits (1) - [`0b2b7fb`](https://github.com/s3fs-fuse/s3fs-fuse/commit/0b2b7fbd92c32fc15f62224c425f74a2acdcfebf) Prevents the etag string buffer from being destroyed during use ### 📊 Changes **3 files changed** (+20 additions, -24 deletions) <details> <summary>View changed files</summary> 📝 `src/fdcache_fdinfo.cpp` (+12 -5) 📝 `src/fdcache_fdinfo.h` (+2 -18) 📝 `src/types.h` (+6 -1) </details> ### 📄 Description ### Relevant Issue (if applicable) n/a ### Details There was a potential bug in the etag string saving process in parallel multipart uploads. Currently, it has not been realized, but it may be a problem in future function addition(planned now), so it has been fixed. This bug is the part that saves etag string when CURL processing is completed in parallel multipart upload. The object has an etag as a pointer, but it is possible that the entity does not exist in the future. Therefore, I added an `etag_entities` member to the `PseudoFdInfo` class and modified it so that the etag entity always exists for the period when the `PseudoFdInfo` object exists. And the `MPPART_INFO` structure has almost the same purpose and member variables as the `filepart` structure, so it is unified as `filepart`. The `MPPART_INFO` changes are useless after this integration, so the above fix and this integration are combined into a single(this) PR. --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>
kerem 2026-03-04 02:03:53 +03:00
Sign in to join this conversation.
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
starred/s3fs-fuse#2134
No description provided.