[coreboot-gerrit] Patch set updated for coreboot: how many K8_REV_F boards with MEM_TRAIN_SEQ==1 are broken?

[Jonathan A. Kollasch (jakllsch@kollasch.net)]

Jonathan A. Kollasch (jakllsch at kollasch.net) just uploaded a new patch set to gerrit, which you can find at http://review.coreboot.org/10916

-gerrit

commit 981b2ac10ef88b921e91dda2fdd1669fcc642b55
Author: Jonathan A. Kollasch <jakllsch at kollasch.net>
Date:   Tue Jul 14 08:20:18 2015 -0500

    how many K8_REV_F boards with MEM_TRAIN_SEQ==1 are broken?
    
    Change-Id: Ib973650f860e8ebf45157b4607780988cbf5647f
    Signed-off-by: Jonathan A. Kollasch <jakllsch at kollasch.net>
---
 src/northbridge/amd/amdk8/raminit_f_dqs.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/src/northbridge/amd/amdk8/raminit_f_dqs.c b/src/northbridge/amd/amdk8/raminit_f_dqs.c
index ee3b609..d5b6bd1 100644
--- a/src/northbridge/amd/amdk8/raminit_f_dqs.c
+++ b/src/northbridge/amd/amdk8/raminit_f_dqs.c
@@ -2074,6 +2074,7 @@ static void train_ram(unsigned nodeid, struct sys_info *sysinfo, struct sys_info
 static inline void train_ram_on_node(unsigned nodeid, unsigned coreid, struct sys_info *sysinfo, unsigned retcall)
 {
 	if(coreid) return; // only do it on core0
+#error this board is broken when multiple sockets populated
 	struct sys_info *sysinfox = (void *)((CONFIG_RAMTOP) - sizeof(*sysinfox));
 	wait_till_sysinfo_in_ram(); // use pci to get it