More info: I'm seeing these 2 entries in console when the gatekeep denies the app. Any idea what this is telling me?
com.apple.message.domain: com.apple.security.assessment.whitelist2
com.apple.message.signature: gslist-555549445c92ec0ac81b3662ba8565553a875a6e
com.apple.message.signature2: 8bb662ddcf832a219e9cd09d016aeb7bd52e5679
com.apple.message.result: fail
com.apple.message.signature3: f0aa56bcede510089225e045db8f8c575a6ddb54
com.apple.message.reason: -67002
com.apple.message.teamid: 72G58AHU7P
SenderMachUUID: D45CB554-D139-30D8-83FC-636CC22ED7D8
assessment denied for gslist
com.apple.message.domain: com.apple.security.assessment.outcome2
com.apple.message.signature2: bundle:UNBUNDLED
com.apple.message.signature3: gslist
com.apple.message.signature5: UNKNOWN
com.apple.message.signature4: 1
com.apple.message.signature: denied:obsolete resource envelope
SenderMachUUID: D45CB554-D139-30D8-83FC-636CC22ED7D8
Post
Replies
Boosts
Views
Activity
Hi Quinn,
Thanks for your comments.
my advice is that you start by doing an end-to-end test with a trivial executable. I was working on that very thing when you replied. Turns out the info.plist must be linked directly into the a.out in order to get past gatekeeper. I found this by trial and error and it would be nice if Apple documented it somewhere.
Adding these arguments to the g++ linker line of my trivial program does that:
Xlinker -sectcreate -Xlinker __TEXT -Xlinker __info_plist -Xlinker </pathTo/Info.plist>
With that change, I can invoke my a.out from the command line in Terminal.app. However I still cannot invoke it by double clicking a.out in Finder. I can live with this limitation.
Next I will this all works with the complete application. I'm skeptical because our app has several shared libs but my trivial test program doesn't.
Norm Green
I still have this problem on Sonoma. Can anyone help with this? I've confirmed launchd runs this plist and sets the new shared memory limits yet they somehow get reset to the lower values.
Nevermind. dlopen() returns a handle not a memory address so not a bug. Sorry for the noise.