]> git.ipfire.org Git - thirdparty/kernel/stable.git/commit
x86/pvh: Call C code via the kernel virtual mapping
authorArd Biesheuvel <ardb@kernel.org>
Wed, 9 Oct 2024 16:04:40 +0000 (18:04 +0200)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Fri, 25 Apr 2025 08:44:01 +0000 (10:44 +0200)
commit32e3456454402967c6ee91074a26dc17bca465cb
treeb491be0e89b8200eff0764892be78e09dde51e77
parent8dfff85d5d7cd616aaeb9cd57c8cda9476e868fe
x86/pvh: Call C code via the kernel virtual mapping

commit e8fbc0d9cab6c1ee6403f42c0991b0c1d5dbc092 upstream.

Calling C code via a different mapping than it was linked at is
problematic, because the compiler assumes that RIP-relative and absolute
symbol references are interchangeable. GCC in particular may use
RIP-relative per-CPU variable references even when not using -fpic.

So call xen_prepare_pvh() via its kernel virtual mapping on x86_64, so
that those RIP-relative references produce the correct values. This
matches the pre-existing behavior for i386, which also invokes
xen_prepare_pvh() via the kernel virtual mapping before invoking
startup_32 with paging disabled again.

Fixes: 7243b93345f7 ("xen/pvh: Bootstrap PVH guest")
Tested-by: Jason Andryuk <jason.andryuk@amd.com>
Reviewed-by: Jason Andryuk <jason.andryuk@amd.com>
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Message-ID: <20241009160438.3884381-8-ardb+git@google.com>
Signed-off-by: Juergen Gross <jgross@suse.com>
[ Stable context update ]
Signed-off-by: Jason Andryuk <jason.andryuk@amd.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
arch/x86/platform/pvh/head.S