[SERVER-42990] Is PCRE 8.43 bug a production concern? Created: 23/Aug/19  Updated: 27/Oct/23  Resolved: 26/Aug/19

Status: Closed
Project: Core Server
Component/s: Build
Affects Version/s: 4.2.0
Fix Version/s: None

Type: Bug Priority: Major - P3
Reporter: James P. Harvey Assignee: Mathew Robinson (Inactive)
Resolution: Works as Designed Votes: 0
Labels: None
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified

Issue Links:
Documented
documents SERVER-47278 Evaluate upgrade of PCRE to 8.44 Closed
Operating System: ALL
Sprint: Dev Tools 2019-09-09
Participants:

 Description   

Apologies for creating a new bug.  This is really a follow-up on "Upgrade PCRE to 8.43" at https://jira.mongodb.org/browse/SERVER-40836

My most recent comment on a different but already closed bug hasn't gotten a response, so I'm not sure if it's just been missed or if like on many bugtrackers that closed bugs never get looked at again.  See https://jira.mongodb.org/browse/SERVER-32773

Anyways, does this PCRE 8.43 bug merely cause a mongodb test to fail?  Should the failure just be ignored?

Or, does this mean that PCRE 8.43 shouldn't be used in production with mongodb?

As the packager of mongodb in the Arch Linux AUR repository, need to know what to do here.  We've always compiled with "--use-system-pcre" and we're running the latest version of PCRE, 8.43.

As half of the "--use-system-X" flags seem to be currently broken and disabled, and with how frequently they change, I'm starting to wonder if any of them are trustworthy, or if we just need to shrug and sigh, and let mongodb pull in all of its dependencies by source.

(Currently using system pcre, snappy, yaml, zlib, stemmer, sasl-client, and tcmalloc.  system wiredtiger, asio, icu, boost, and surprisingly mongo-c are broken.)



 Comments   
Comment by Mathew Robinson (Inactive) [ 26/Aug/19 ]

Hey jamespharvey20,

I hope the above answers your question. Since we're waiting for PCRE 8.44 to do an upgrade I'm going to close this as Works as Designed.

As the Arch packager I leave it up to you to decide if the one edge case that can cause a crash with PCRE 8.43 is worth using the system library or not.

Comment by Mathew Robinson (Inactive) [ 26/Aug/19 ]

If PCRE 8.43 is used without the fix then the (*LF) prefix option in a $regex query would cause the database to throw a DBException. I would consider this a production problem, but I have no way to know if that prefix option is widely used.

Comment by Andrew Morrow (Inactive) [ 23/Aug/19 ]

jamespharvey20 - Regarding your mention of a comment on SERVER-32773 - I'd generally advise against leaving follow-up comments on a closed ticket, except if you think that the issue was not actually fixed and that it should be re-opened. In general, it is better to open new tickets - there is a system in place for triage of new tickets, but not so much for comments on closed. Also, for things that are more like a discussion or set of questions (say, regarding the issues around the --use-system- flags that you and I have talked about a few times), I'd also recommend starting out with an email to the moderated group https://groups.google.com/forum/#!forum/mongodb-dev. That way, we can have a conversation around the issue before deciding on if tickets should be filed.

Generated at Thu Feb 08 05:01:58 UTC 2024 using Jira 9.7.1#970001-sha1:2222b88b221c4928ef0de3161136cc90c8356a66.