Add an initial version of some tools to compare (extended) K8 memory settings.

This generates (dirty) html with interpreted differences between PCI dumps,
based on the K8 socket F bkdg.

Signed-off-by: Ward Vandewege <ward@gnu.org>
Acked-by: Stepan Reinauer <stepan@coresystems.de>



git-svn-id: svn://svn.coreboot.org/coreboot/trunk@4886 2b7e53f0-3cfb-0310-b3e9-8179ed1497e1

1 files changed, 32 insertions, 0 deletions

diff --git a/util/amdtools/README b/util/amdtools/README
new file mode 100644
index 0000000000..a1601fe8c7
--- /dev/null
+++ b/util/amdtools/README
@@ -0,0 +1,32 @@
+
+
+This is a set of tools to compare (extended) K8 memory settings.
+
+Before you can use them, you need to massage the relevant BKDG sections into
+useable data. Here's how.
+
+First, you need to acquire a copy of the K8 BKDG. Go here:
+
+  Rev F: http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/32559.pdf
+
+Then make sure pdftotext is installed (it's in the poppler-utils package on Debian/Ubuntu).
+
+Now run the bkdg through pdftotext:
+
+  pdftotext -layout 32559.pdf 32559.txt
+
+Now extract sections 4.5.15 - 4.5.19 from the file, and save it separately, say as bkdg-raw.data.
+
+Finally run the txt file through the parse-bkdg.pl script like so:
+
+  parse-bkdg.pl < bkdg-raw.data > bkdg.data
+
+Now we have the bkdg.data file that is used by the other scripts.
+
+If you want to test the scripts without doing all this work, you can use some
+sample input files from the 'example_input/' directory. 
+
+-- 
+Ward Vandewege, 2009-10-28.
+ward@jhvc.com
+