| FazBrowse GitHub Viewer | Trending | | Home |
| Tools: [Download Repo ZIP] [Original HTTPS Page] |
|
Correct for ARM64, not so much for other ARM. We test (and build) on Wheezy for ARMv6 and ARMv7, so that gives us glibc >= 2.13 for both of those. When you include kernel version things get even more complicated because the lowest we're running for ARMv7 is 3.2.34 but on ARMv6 we get 4.1.19 because we're using a Jessie boot setup to get the best out of our Pi's. I really don't know that we want to be that specific here. I've argued against including kernel versions in here previously since it rarely matters when you account for libc and just makes things messy and complicated. I'm willing to bet that exactly zero people have found the kernel version we list helpful in any practical way. |
Sorry, something went wrong.
|
I guess this list is actually difficult to properly maintain. Is there any way we could automate something that produces a list of supported systems and their minimum build versions? |
Sorry, something went wrong.
|
@BridgeAR I guess an Ansible script or two to run across all our release- machines would get a pretty good picture? While we're at it the same thing could be run across all of test- to get a snapshot of what's actually in use. |
Sorry, something went wrong.
I think test- is probably more valuable than release- right? We don't actually check that node runs on the machines we build it on. I guess in some cases we might have "experimental" or Tier 2 platforms that we build for but don't run CI on? |
Sorry, something went wrong.
Yeah, I figured that. My vote would be for switching to a set of "these are the platforms we build and test on", but it seems wrong to leave the arm reference wrong. So how about I leave ARM as it was, and just change ARM64? |
Sorry, something went wrong.
|
yeah, put a lid back on that can of worms |
Sorry, something went wrong.
Sorry, something went wrong.
There was a problem hiding this comment.
LGTM
Sorry, something went wrong.
Sorry, something went wrong.
This is already true in practice. PR-URL: nodejs#19164 Fixes: nodejs/build#1164 Reviewed-By: Rod Vagg <rod@vagg.org> Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
This is already true in practice. PR-URL: #19164 Fixes: nodejs/build#1164 Reviewed-By: Rod Vagg <rod@vagg.org> Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
This is already true in practice. PR-URL: #19164 Fixes: nodejs/build#1164 Reviewed-By: Rod Vagg <rod@vagg.org> Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
This is already true in practice. PR-URL: nodejs#19164 Fixes: nodejs/build#1164 Reviewed-By: Rod Vagg <rod@vagg.org> Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
|
should this be backported to 8.x? If so, a separate backport PR is needed. |
Sorry, something went wrong.
This is already true in practice. Backport-PR-URL: #23275 PR-URL: #19164 Fixes: nodejs/build#1164 Reviewed-By: Rod Vagg <rod@vagg.org> Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
| Back | FazBrowse Home | New Git URL |
This is already true in practice.
Fixes: nodejs/build#1164
cc/ @nodejs/build @rvagg @gdams
@rvagg I'm not sure if this applies to arm (not arm64) as well.