Replies: 2 comments 1 reply
|
The default configuration sets the arch to core2, unless the formula defines that the software has runtime detection of the CPU features. Did you run into issues without defining the arch? Or did you set it out of an abundance of caution? |
1 reply
|
You can use Or, you can just install the bottles (i.e. don't build from source) which should just work on your various machines. |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Additional Info
Output of
brew configOutput of
brew doctorDescription of issue
I see that in the 7.0 release the config HOMEBREW_ARCH is listed as deprecated and it will behave as if you used 'native' in the future. This means that any packages that are compiled will used the latest architecture features of the machine doing the building. I did search for a reason for this change, but I failed to find a place where it was discussed and the reason given.
I use homebrew in an environment with a large mixture of x86 machines with different processors and put the HOMEBREW_PREFIX on a shared NFS disk so multiple people can use it. Without HOMEBREW_ARCH, I would need to do all builds on the oldest machine to make sure these packages can be used by all clients.
Yes, I understand this isn't really a "supported" configuration, but it is useful when you have to deal with a work environment where the standard "stable" Linux configuration is very old and doesn't include the tools users want to use.
All reactions