docs/hcl/x200.html: Notes about raminit, S3 and microcode

2 files changed, 47 insertions, 0 deletions

diff --git a/docs/hcl/x200.html b/docs/hcl/x200.html
index 8567a91c..b8bbfd9f 100644
--- a/docs/hcl/x200.html
+++ b/docs/hcl/x200.html
@@ -307,6 +307,48 @@
 
 	<div class="section">
 
+		<h1 id="ram_s3_microcode">RAM, S3 and microcode updates</h1>
+
+			<p>
+				Not all memory modules work. Most of the default ones do, but you have to be careful
+				when upgrading to 8GiB; some modules work, some don't.
+			</p>
+
+			<p>
+				pehjota started collecting some steppings for different CPUs on several X200 laptops.
+				You can get the CPUID by running: <br/>
+				$ <b>dmesg | sed -n 's/^.* microcode: CPU0 sig=0x\([^,]*\),.*$/\1/p'</b>
+			</p>
+
+			<p>
+				What pehjota wrote:
+				 The laptops that have issues resuming from suspend, as well as a laptop that (as I mentioned earlier in #libreboot) won't boot with any Samsung DIMMs, all have CPUID 0x10676 (stepping M0).
+			</p>
+
+			<p>
+				What pehjota wrote:
+				Laptops with CPUID 0x167A (stepping R0) resume properly every time and work with Samsung DIMMs.  I'll 
+				need to do more testing on more units to better confirm these trends, but it looks like the M0 microcode 
+				is very buggy.  That would also explain why I didn't have issues with Samsung DIMMs with the Lenovo BIOS 
+				(which would have microcode updates).  I wonder if VT-x works on R0.
+			</p>
+
+			<p>
+				What pehjota wrote:
+				As I said, 10676 is M0 and 1067A is R0; those are the two CPUIDs and steppings for Intel Core 2 Duo P8xxx CPUs with factory microcode.  (1067 is the family and model, and 6 or A is the stepping ID.)
+			</p>
+
+			<p>
+				<b>
+					TODO: check the CPUIDs and test S3 resume and/or KVM on any C2D systems (including non-P8xxx ones, which I don't have here) you have available.  I'd be curious if you could confirm these results.
+				</b>
+				It might not be coreboot that's buggy with raminit/S3; it might just be down to the microcode updates.
+			</p>
+
+	</div>
+
+	<div class="section">
+
 		<h2 id="unsorted">Unsorted notes</h2>
 
 <pre>
diff --git a/docs/tasks.html b/docs/tasks.html
index 13961ec1..703fb90a 100644
--- a/docs/tasks.html
+++ b/docs/tasks.html
@@ -119,6 +119,11 @@
 		<h1>Platform-specific bugs</h1>
 			<ul>
 				<li>
+					GM45: investigate S3/raminit on all models (CPU stepping/cpuid).
+					See <a href="hcl/x200.html#ram_s3_microcode">hcl/x200.html#ram_s3_microcode</h1>
+
+				</li>
+				<li>
 					all thinkpads: When the system is running, plugging in an ethernet cable
 					doesn't always work (no network), you have to try several times. Booting with
 					an ethernet cable attached is reliable. Debug this and fix it.