Bug 239653
| Summary: | databases/mariadb103-server upgrade to 10.3.17 | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | Ports & Packages | Reporter: | Michael Glaus <michael.glaus> | ||||||
| Component: | Individual Port(s) | Assignee: | Bernard Spil <brnrd> | ||||||
| Status: | Closed FIXED | ||||||||
| Severity: | Affects Many People | CC: | brnrd, d.stoye, linus.sundqvist, pascal.christen, santhosh.raju | ||||||
| Priority: | --- | Keywords: | security | ||||||
| Version: | Latest | Flags: | bugzilla:
maintainer-feedback?
(brnrd) |
||||||
| Hardware: | Any | ||||||||
| OS: | Any | ||||||||
| Attachments: |
|
||||||||
|
Description
Michael Glaus
2019-08-05 07:58:09 UTC
Created attachment 206288 [details]
Patch for update
I've built it when it was released (all options enabled) and it failed on some C error. Can you please try that and see if you can come up with a patch? (In reply to Bernard Spil from comment #2) I didn't get any errors built with FB112 and FB113. Can you give me the error you got so I can take a look at it? Something related to this? https://github.com/openbsd/ports/commit/ef7c6a78f5a869f65048018fc84eaeb10b3fca6b (In reply to Pascal Christen from comment #3) That patch from Stephen Henderson only modifies some linker flags. Issues that I'm running into (all options enabled!) are more like https://jira.mariadb.org/browse/MDEV-20258 https://jira.mariadb.org/browse/MDEV-20248 Help much appreciated. Work is crazy at the moment! Created attachment 206461 [details] Fixes build failures for MDEV-20248 (In reply to Bernard Spil from comment #5) Hello Bernard In addition to the above patch for update submitted by Michael, I have an additional patch that defines "JEMALLOC_USABLE_SIZE_CONST" to be "const" in the file indicated by the build failure in https://jira.mariadb.org/browse/MDEV-20248 I have patched the file with the following lines #ifdef __FreeBSD__ #define JEMALLOC_USABLE_SIZE_CONST const #endif this is based on my look up of the function prototype of "malloc_usable_size()" from jemalloc(3) man page. (I hope the fix is correct.) The seems to solve the build failure issue when "all options" are selected like you have mentioned in your reply. I am not able to replicate the other Bug report https://jira.mariadb.org/browse/MDEV-20258 I shall try out the mutually exclusive options at the end and see if I can replicate the build failure from my end. Comment on attachment 206461 [details]
Fixes build failures for MDEV-20248
Thanks!
Patch looks good to me.
Adding another patch for the jemalloc issue, marking TokuDB BROKEN for now.
Running builds on 120/112 amd64/i386, expect a commit shortly!
(In reply to Bernard Spil from comment #7) Did the builds go fine? any news? as this is a security related update, is there still something todo, or can it be released? (In reply to Daniel from comment #9) Got pushed a few days ago: https://www.freshports.org/commit.php?category=databases&port=mariadb103-server&files=yes&message_id=201909011557.x81Fv2e0086111@repo.freebsd.org thx, i'm sorry Forgot PR reference in commit msg :/ |
