]> git.proxmox.com Git - mirror_edk2.git/commit - OvmfPkg/OvmfPkg.dec
OvmfPkg: Sec: assert the build-time calculated end of the scratch buffer
authorLaszlo Ersek <lersek@redhat.com>
Mon, 30 Nov 2015 18:41:20 +0000 (18:41 +0000)
committerlersek <lersek@Edk2>
Mon, 30 Nov 2015 18:41:20 +0000 (18:41 +0000)
commit9beac0d847bf9c299fe6c05b0fe7041a75bffa67
treea8853eb7d340a3bb30c61f398aafa6ab48da1cb9
parent320b4f084a256e16a7f0cbfa5f0b5c8e0bb1a0ac
OvmfPkg: Sec: assert the build-time calculated end of the scratch buffer

The DecompressMemFvs() function in "OvmfPkg/Sec/SecMain.c" uses more
memory, temporarily, than what PEIFV and DXEFV will ultimately need.
First, it uses an output buffer for decompression, second, the
decompression itself needs a scratch buffer (and this scratch buffer is
the highest area that SEC uses).

DecompressMemFvs() used to be called on normal boots only (ie. not on S3
resume), which is why the decompression output buffer and the scratch
buffer were allowed to scribble over RAM. However, we'll soon start to
worry during S3 resume that the runtime OS might tamper with the
pre-decompressed PEIFV, and we'll decompress the firmware volumes on S3
resume too, from pristine flash. For this we'll need to know the end of
the scratch buffer in advance, so we can prepare a non-malicious OS for
it.

Calculate the end of the scratch buffer statically in the FDF files, and
assert in DecompressMemFvs() that the runtime decompression will match it.

Contributed-under: TianoCore Contribution Agreement 1.0
Signed-off-by: Laszlo Ersek <lersek@redhat.com>
Reviewed-by: Jordan Justen <jordan.l.justen@intel.com>
git-svn-id: https://svn.code.sf.net/p/edk2/code/trunk/edk2@19036 6f19259b-4bc3-4df7-8a09-765794883524
OvmfPkg/DecomprScratchEnd.fdf.inc [new file with mode: 0644]
OvmfPkg/OvmfPkg.dec
OvmfPkg/OvmfPkg.fdf.inc
OvmfPkg/OvmfPkgIa32.fdf
OvmfPkg/OvmfPkgIa32X64.fdf
OvmfPkg/OvmfPkgX64.fdf
OvmfPkg/Sec/SecMain.c
OvmfPkg/Sec/SecMain.inf