Page MenuHomeFreeBSD
Differential D58357

nvme: Explicitly cast caddr_t values
ClosedPublic
Actions

Authored by des on Jul 20 2026, 1:18 PM.
Tags
None
Referenced Files
Unknown Object (File)
Fri, Aug 21, 4:54 AM
Unknown Object (File)
Sun, Aug 16, 5:38 PM
Unknown Object (File)
Thu, Aug 13, 12:29 PM
Unknown Object (File)
Tue, Aug 11, 11:41 AM
Unknown Object (File)
Tue, Aug 11, 11:16 AM
Unknown Object (File)
Tue, Aug 11, 9:29 AM
Unknown Object (File)
Sun, Aug 9, 6:22 AM
Unknown Object (File)
Sat, Aug 8, 11:05 PM
View All 44 Files
Subscribers
dab

Details

Reviewers
imp
asomers
bnovkov
Group Reviewers
Klara
Commits
rGc4d8b894dd5e: nvme: Explicitly cast caddr_t values
rG3f99b452687c: nvme: Explicitly cast caddr_t values
rGbd30d1ad78e1: nvme: Explicitly cast caddr_t values
Summary

Sponsored by: Klara, Inc.
Sponsored by: NetApp, Inc.
MFC after: 1 week
Fixes: 6d0001d44490 ("nvme: add support for DIOCGIDENT")

Diff Detail

Repository
rG FreeBSD src repository
Lint
Lint Not Applicable
Unit
Tests Not Applicable

Event Timeline

des created this revision.Jul 20 2026, 1:18 PM
Herald added a subscriber: dab. · View Herald TranscriptJul 20 2026, 1:18 PM
des requested review of this revision.Jul 20 2026, 1:18 PM
Harbormaster completed remote builds in B74953: Diff 182289.Jul 20 2026, 1:18 PM
imp added a comment.Jul 20 2026, 1:59 PM

Why bother?

des added a comment.Jul 20 2026, 2:19 PM
In D58357#1338417, @imp wrote:

Why bother?

It's causing a build failure downstream.

bnovkov accepted this revision.Jul 20 2026, 9:54 PM
This revision is now accepted and ready to land.Jul 20 2026, 9:54 PM
imp accepted this revision.Jul 20 2026, 10:00 PM
In D58357#1338418, @des wrote:
In D58357#1338417, @imp wrote:

Why bother?

It's causing a build failure downstream.

What's the failure. the change is fine (I should have hit accept the first time), but I was hoping to learn something since caddr_t and uint8_t* should be the same (modulo signedness... is that causing the issue or something else?).

des added a comment.Jul 21 2026, 8:02 AM
In D58357#1338639, @imp wrote:
In D58357#1338418, @des wrote:

It's causing a build failure downstream.

What's the failure. the change is fine (I should have hit accept the first time), but I was hoping to learn something since caddr_t and uint8_t* should be the same (modulo signedness... is that causing the issue or something else?).

The code is being built in an environment where caddr_t is defined as intptr_t rather than char * (still allows pointer arithmetic but prevents dereferencing). Every other assignment of a caddr_t value in the nvme driver already has the correct cast in place, this one didn't.

Closed by commit rGbd30d1ad78e1: nvme: Explicitly cast caddr_t values (authored by des). · Explain WhyJul 21 2026, 11:34 AM
This revision was automatically updated to reflect the committed changes.
des added a commit: rGbd30d1ad78e1: nvme: Explicitly cast caddr_t values.
des added a commit: rG3f99b452687c: nvme: Explicitly cast caddr_t values.Tue, Jul 28, 9:22 AM
des added a commit: rGc4d8b894dd5e: nvme: Explicitly cast caddr_t values.

Revision Contents
Changeset List

PathSize
sys/
dev/
nvme/
2 lines

Diff 182376

View Options

sys/dev/nvme/nvme_ns.c

Loading...