Opened 2 years ago

Closed 15 months ago

Last modified 10 months ago

#63811 closed defect (fixed)

acpica @20210930: ld: unaligned pointer(s) for architecture arm64

Reported by: Ranguvar (Devin Cofer) Owned by: kencu (Ken)
Priority: Normal Milestone:
Component: ports Version: 2.7.1
Keywords: arm64 Cc: i0ntempest
Port: acpica

Description

M1 Max MacBook Pro running macOS Monterey 12.0.1

Build log attached

Attachments (1)

acpica-main.log (361.3 KB) - added by Ranguvar (Devin Cofer) 2 years ago.
build log

Download all attachments as: .zip

Change History (7)

Changed 2 years ago by Ranguvar (Devin Cofer)

Attachment: acpica-main.log added

build log

comment:1 Changed 2 years ago by Ranguvar (Devin Cofer)

Keywords: monterey added; monteray removed

comment:2 Changed 15 months ago by kencu (Ken)

Keywords: monterey removed

not specific to Monterey, happens on all systems when built arm64

comment:3 Changed 15 months ago by kencu (Ken)

Owner: set to kencu
Resolution: fixed
Status: newclosed

In 87b92651bdefd4fc6cbf2943ee957c23cefd3a9d/macports-ports (master):

acpica: update to 20221022

moved to github

set supported_archs... this is Intel-specific software

closes: #63811

comment:4 Changed 10 months ago by ryandesign (Ryan Carsten Schmidt)

Cc: i0ntempest added

arm64 (and PowerPC...) builds were re-enabled in [ce28396921980998935ee1b3025fbb42a75ae328/macports-ports] and [3de9eefec887788181c3197208c6bd9f487a3934/macports-ports] by forcing macOS 11 deployment target; apparently macOS 12 deployment target and later is what complains about these unaligned pointers.

Here's how another project fixed it though: https://github.com/RfidResearchGroup/proxmark3/issues/1485#issuecomment-982089666

comment:5 Changed 10 months ago by ryandesign (Ryan Carsten Schmidt)

Here is the acpica bug report about this problem: https://github.com/acpica/acpica/issues/781

comment:6 Changed 10 months ago by kencu (Ken)

some info here:

https://developer.apple.com/forums/thread/651200

probably here:

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=51628

https://stackoverflow.com/questions/68640690/arm-gcc-address-of-packed-struct-warning

all this makes me wonder if forcing the deployment target to 11 to allow this to pass through the linker is actually doing the right thing, or is it just getting past the warnings/errors but emitting broken code?

Note: See TracTickets for help on using tickets.