You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The OS currently fails to build with a proper multiboot section on ReleaseFast and Small builds.
For the header to be valid, the entire multiboot header must apparently be in the first 8KB of the OS image, and also be 32-bit aligned. (And ideally should be included as early as possible)
As of 0.8.0-dev.2193+fe14270f4, building in Debug and Safe seems to put the header in the 4096th byte. Building as Fast, Small, and Safe without the panic function seems to place the header at the 8192th byte.
It seems like removing the ALIGN(4K) for .text from the linker.ld makes the image valid on Fast, Small, and Safe + panic.
This changes the location of the header for the following build modes:
Debug/Safe: No change, 0x1000 (or 4096)
Fast: 0x1030
Small: 0x1024
Safe without the panic function: 0x1030
Since I don't really get the reason behind this - and hence I'm not sure what the best way to deal with the problem - I'm not going to make a PR. Well, at least for now.
The text was updated successfully, but these errors were encountered:
The OS currently fails to build with a proper multiboot section on ReleaseFast and Small builds.
For the header to be valid, the entire multiboot header must apparently be in the first 8KB of the OS image, and also be 32-bit aligned. (And ideally should be included as early as possible)
As of
0.8.0-dev.2193+fe14270f4
, building in Debug and Safe seems to put the header in the 4096th byte. Building as Fast, Small, and Safe without thepanic
function seems to place the header at the 8192th byte.It seems like removing the
ALIGN(4K)
for.text
from thelinker.ld
makes the image valid on Fast, Small, and Safe +panic
.This changes the location of the header for the following build modes:
0x1000
(or 4096)0x1030
0x1024
panic
function:0x1030
Since I don't really get the reason behind this - and hence I'm not sure what the best way to deal with the problem - I'm not going to make a PR. Well, at least for now.
The text was updated successfully, but these errors were encountered: