Rudolf Marek wrote:
>> -    /* ROM memory cycles go to LPC. */
>> +    /* Only ROM memory cycles go to LPC. */
>
> I think it should be,
>
> > +   /* Only memory ROM cycles go to LPC.  */
>
> Because this bit is telling if all memory cycles should go to LPC or just 
> those from ROM range. Please try to rephase that.

Updated patch:

Comment changes, add pointer to PCIe bridge documentation.

Signed-off-by: Tobias Diedrich <ranma+coreb...@tdiedrich.de>

---

Index: src/southbridge/via/k8t890/k8t890_ctrl.c
===================================================================
--- src/southbridge/via/k8t890/k8t890_ctrl.c.orig       2010-11-03 
14:53:07.000000000 +0100
+++ src/southbridge/via/k8t890/k8t890_ctrl.c    2010-11-03 15:51:23.000000000 
+0100
@@ -154,7 +154,11 @@
 
        pci_write_config8(dev, 0x47, 0x30);
 
-       /* VT8237R specific configuration  other SB are done in their own 
directories */
+       /*
+        * VT8237R specific configuration,
+        * other SB are done in their own directories:
+        * VT8237A and VT8237S are handled in vt8237_ctrl.c
+        */
 
        device_t devsb = dev_find_device(PCI_VENDOR_ID_VIA,
                                         PCI_DEVICE_ID_VIA_VT8237R_LPC, 0);
Index: src/southbridge/via/vt8237r/vt8237r_lpc.c
===================================================================
--- src/southbridge/via/vt8237r/vt8237r_lpc.c.orig      2010-11-03 
15:50:51.000000000 +0100
+++ src/southbridge/via/vt8237r/vt8237r_lpc.c   2010-11-03 16:02:37.000000000 
+0100
@@ -457,7 +457,18 @@
        /* I/O recovery time, default IDE routing */
        pci_write_config8(dev, 0x4c, 0x04);
 
-       /* ROM memory cycles go to LPC. */
+       /*
+        * Bit | Meaning
+        *   7 | 1: Only ROM memory cycles go to LPC instead of all memory
+        *     |    cycles.
+        *   6 | 0: Internal ISA cycles do not arbitrate with secondary IDE
+        *   5 | 0: Disable LPC RTC
+        *   4 | 0: Disable LPC Keyboard
+        *   3 | 0: Disable Port 0x62/0x66 to LPC
+        *   2 | 0: Disable Port 0x62/0x66 MCCS# chipselect decoding
+        *   1 | 0: Disable A20M# signal (signal not asserted)
+        *   0 | 0: Disable NMI on PCI parity error
+        */
        pci_write_config8(dev, 0x59, 0x80);
 
        /*
@@ -479,7 +490,18 @@
        /* I/O recovery time, default IDE routing */
        pci_write_config8(dev, 0x4c, 0x44);
 
-       /* ROM memory cycles go to LPC. */
+       /*
+        * Bit | Meaning
+        *   7 | 1: Only ROM memory cycles go to LPC instead of all memory
+        *     |    cycles.
+        *   6 | 0: Internal ISA cycles do not arbitrate with secondary IDE
+        *   5 | 0: Disable LPC RTC
+        *   4 | 0: Disable LPC Keyboard
+        *   3 | 0: Disable Port 0x62/0x66 to LPC
+        *   2 | 0: Disable Port 0x62/0x66 MCCS# chipselect decoding
+        *   1 | 0: Disable A20M# signal (signal not asserted)
+        *   0 | 0: Disable NMI on PCI parity error
+        */
        pci_write_config8(dev, 0x59, 0x80);
 
        /*
Index: src/southbridge/via/k8t890/k8t890_pcie.c
===================================================================
--- src/southbridge/via/k8t890/k8t890_pcie.c.orig       2010-11-03 
14:53:07.000000000 +0100
+++ src/southbridge/via/k8t890/k8t890_pcie.c    2010-11-03 15:51:23.000000000 
+0100
@@ -24,6 +24,12 @@
 #include <device/pci_ids.h>
 #include "k8t890.h"
 
+/*
+ * Note:
+ * The pcie bridges are similar to the VX800 ones documented at
+ * http://linux.via.com.tw/
+ */
+
 static void peg_init(struct device *dev)
 {
        u8 reg;


-- 
coreboot mailing list: coreboot@coreboot.org
http://www.coreboot.org/mailman/listinfo/coreboot

Reply via email to