summaryrefslogtreecommitdiff
path: root/tools/bddb/defs.php
diff options
context:
space:
mode:
authorwdenk <wdenk>2003-06-19 23:40:20 +0000
committerwdenk <wdenk>2003-06-19 23:40:20 +0000
commit6dd652fa4d8591a32e2707a91f4582ed13011b17 (patch)
tree52d7c375a3a33d634e7c30908a75bf9616d92854 /tools/bddb/defs.php
parent52f52c1494eedaeacccad6e2331f4f638b48f5ab (diff)
Patches by Murray Jensen, 17 Jun 2003:
- Hymod board database mods: add "who" field and new xilinx chip types - provide new "init_cmd_timeout()" function so code external to "common/main.c" can use the "reset_cmd_timeout()" function before entering the main loop - add DTT support for adm1021 (new file dtt/adm1021.c; config slightly different. see include/configs/hymod.h for an example (requires CONFIG_DTT_ADM1021, CONFIG_DTT_SENSORS, and CFG_DTT_ADM1021 defined) - add new "eeprom_probe()" function which has similar args and behaves in a similar way to "eeprom_read()" etc. - add 8260 FCC ethernet loopback code (new "eth_loopback_test()" function which is enabled by defining CONFIG_ETHER_LOOPBACK_TEST) - gdbtools copyright update - ensure that set_msr() executes the "sync" and "isync" instructions after the "mtmsr" instruction in cpu/mpc8260/interrupts.c - 8260 I/O ports fix: Open Drain should be set last when configuring - add SIU IRQ defines for 8260 - allow LDSCRIPT override and OBJCFLAGS initialization: change to config.mk to allow board configurations to override the GNU linker script, selected via the LDSCRIPT, make variable, and to give an initial value to the OBJCFLAGS make variable - 8260 i2c enhancement: o correctly extends the timeout depending on the size of all queued messages for both transmit and receive o will not continue with receive if transmit times out o ensures that the error callback is done for all queued tx and rx messages o correctly detects both tx and rx timeouts, only delivers one to the callback, and does not overwrite an earlier error o logic in i2c_probe now correct - add "vprintf()" function so that "panic()" function can be technically correct - many Hymod board changes
Diffstat (limited to 'tools/bddb/defs.php')
-rw-r--r--tools/bddb/defs.php12
1 files changed, 9 insertions, 3 deletions
diff --git a/tools/bddb/defs.php b/tools/bddb/defs.php
index 0393dbda6..9361419d4 100644
--- a/tools/bddb/defs.php
+++ b/tools/bddb/defs.php
@@ -58,8 +58,8 @@
$zbt_nbits = array(0,19,20,21,22);
// Xilinx attributes
- $xlxtyp_vals = array('','XCV300E','XCV400E','XCV600E');
- $xlxspd_vals = array('','6','7','8');
+ $xlxtyp_vals = array('','XCV300E','XCV400E','XCV600E','XC2V2000','XC2V3000','XC2V4000','XC2V6000');
+ $xlxspd_vals = array('','6','7','8','4','5');
$xlxtmp_vals = array('','COM','IND');
$xlxgrd_vals = array('','NORMAL','ENGSAMP');
@@ -449,7 +449,13 @@
// generate a (possibly not unique) random vendor ethernet address
// (setting bit 6 in the ethernet address - motorola wise i.e. bit 0
// is the most significant bit - means it is not an assigned ethernet
- // address). Also, make sure it is NOT a multicast ethernet address.
+ // address - it is a "locally administered" address). Also, make sure
+ // it is NOT a multicast ethernet address (by setting bit 7 to 0).
+ // e.g. the first byte of all ethernet addresses generated here will
+ // have 2 in the bottom two bits (incidentally, these are the first
+ // two bits transmitted on the wire, since the octets in ethernet
+ // addresses are transmitted LSB first).
+
function gen_eth_addr($serno) {
$ethaddr_high = (mt_rand(0, 65535) & 0xfeff) | 0x0200;