Fix for bug where the iniatialisation of the MAC data private strucure
is copied into the wrong location. More details in the ticket #9547
Simple fix, but good to work through the patching and submission process.
I'll try follow up with a small refactoring of the code in that area.
Tested on an arv4518pw.

Signed-off-by: Conor O'Gorman <conor at dna.ie>

---

Index: target/linux/lantiq/patches-2.6.39/200-mach-arv45xx.patch
===================================================================
--- target/linux/lantiq/patches-2.6.39/200-mach-arv45xx.patch   (revision 27170)
+++ target/linux/lantiq/patches-2.6.39/200-mach-arv45xx.patch   (working copy)
@@ -276,7 +276,7 @@
 +arv45xx_register_ethernet(void)
 +{
 +#define ARV45XX_BRN_MAC                       0x3f0016
-+      memcpy_fromio(&ltq_eth_data.mac,
++      memcpy_fromio(&ltq_eth_data.mac.sa_data,
 +              (void *)KSEG1ADDR(LTQ_FLASH_START + ARV45XX_BRN_MAC), 6);
 +      ltq_register_etop(&ltq_eth_data);
 +}
@@ -285,7 +285,7 @@
 +arv75xx_register_ethernet(void)
 +{
 +#define ARV75XX_BRN_MAC                       0x7f0016
-+      memcpy_fromio(&ltq_eth_data.mac,
++      memcpy_fromio(&ltq_eth_data.mac.sa_data,
 +              (void *)KSEG1ADDR(LTQ_FLASH_START + ARV75XX_BRN_MAC), 6);
 +      ltq_register_etop(&ltq_eth_data);
 +}
@@ -294,7 +294,7 @@
 +bewan_register_ethernet(void)
 +{
 +#define BEWAN_BRN_MAC                 0x3f0014
-+      memcpy_fromio(&ltq_eth_data.mac,
++      memcpy_fromio(&ltq_eth_data.mac.sa_data,
 +              (void *)KSEG1ADDR(LTQ_FLASH_START + BEWAN_BRN_MAC), 6);
 +      ltq_register_etop(&ltq_eth_data);
 +}
_______________________________________________
openwrt-devel mailing list
openwrt-devel@lists.openwrt.org
https://lists.openwrt.org/mailman/listinfo/openwrt-devel

Reply via email to