-
Notifications
You must be signed in to change notification settings - Fork 2.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
nginx aarch64-musl cannot serve over HTTPS #30945
Comments
Can confirm, this also happens on aarch64 glibc |
3 weeks ago I had HTTPS working on my RPI3B (aarch64-musl), I can not test it now. The only difference was that I was using a local CA certificate. |
This only affects http2, and seems to be a cross compilation issue, when compiled natively it seems to work on aarch64 (glibc). @cyckl can you check if your config works without http2? |
https://github.com/void-linux/void-packages/blob/master/srcpkgs/nginx/template#L88 this is most likely the issue, we use a ton of wrong values for aarch64 in there. |
This worked perfectly. Thank you so much! For future reference I remember turning off http2 without any luck... |
System
It's a Raspberry Pi 4B
Void 5.4.83_1 aarch64-musl Unknown uptodate rF
nginx-1.18.0_4
Custom image built with
rpi4-kernel
andrpi4-base
Expected behavior
After serving a site on SSL, nginx should send correct response to request for HTTPS content
The following example was taken from
curl
output with the following conditions:Void 5.11.18_1 x86_64 GenuineIntel uptodate rFFF
nginx-1.18.0_4
Actual behavior
nginx will send an empty response on SSL requests, at least on my aarch64-musl install of the package
The following example was taken from
curl
output with the following conditions:Void 5.4.83_1 aarch64-musl Unknown uptodate rF
nginx-1.18.0_4
Steps to reproduce the behavior
curl
from websiteExtra information
error.log
located in/var/log/nginx/error.log
Ideas
As I've tested the exact same package on different architectures, the only difference between the two is that there is a patch for the nginx package applied for ARM systems. It appears that the patch is in order to adjust certain configuration values on compilation to fit the ARM architecture, which also leads me to believe that it could be related. A good way to test would be with a few different architectures to see if it's only limited to ARM or maybe something platform specific.
It could also be musl specific, as that is another difference between the two systems and that difference has been known to break packages historically.
The text was updated successfully, but these errors were encountered: