Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Hi!

No. The hardware modded WL-500gP with 128 MB RAM has the same problem with kamikaze 2.6, as the RT-N16.

vargalex

codertux wrote:

this discussion is deviating from the rt-n16 to the wl-500gp I see...

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Yes, I've downloaded it, but have not flashed it by now, i really need to build a serial console first so that i can check the output msg come from cfe. any report will be feedbacked here as the test finished.

yellowbug

vargalex wrote:

Hi Yellowbug!

Please check the firmware in #45 (https://forum.openwrt.org/viewtopic.php?pid=102330#p102330) for WL-500gP!

vargalex

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

vargalex wrote:

Hi!

No. The hardware modded WL-500gP with 128 MB RAM has the same problem with kamikaze 2.6, as the RT-N16.

vargalex

Well, not necesseraly as Andrei showed at https://dev.openwrt.org/ticket/6580, that the 2.4 branch of Openwrt also fails on the RT-N16, which suggests, that something else (also) causes a boot failure (perhaps additionaly to the 128Mb access problem).
(The prreviously suggested NVRAM settings are questionable on the RT-N16, since the RT-N16 boots with other 2.6 firmwares)

54 (edited by vargalex 2010-02-11 08:33:34)

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Hi!

This is not so simple. By me runs the 2.4 branch, and the own patched 2.6. But by others (with 128MB modded WL-500gP too) fails my own buildet kamikaze with kernel 2.6. From 2.4 branch by others (with 128MB modded WL-500gP) I have no information. So maybe the 2.4 will not run by others too... So the 2.4 should be tested first on other 128MB modded WL-500gP.

vargalex

umiki wrote:
vargalex wrote:

Hi!

No. The hardware modded WL-500gP with 128 MB RAM has the same problem with kamikaze 2.6, as the RT-N16.

vargalex

Well, not necesseraly as Andrei showed at https://dev.openwrt.org/ticket/6580, that the 2.4 branch of Openwrt also fails on the RT-N16, which suggests, that something else (also) causes a boot failure (perhaps additionaly to the 128Mb access problem).
(The prreviously suggested NVRAM settings are questionable on the RT-N16, since the RT-N16 boots with other 2.6 firmwares)

55 (edited by poisonhzkj 2010-02-12 12:16:28)

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

I my case, with wl500gp 128m ram mod.

1. Build a firmware from trunk
2. In build_dir manually change the build_dir/toolchain-mipsel_gcc-4.3.3+cs_uClibc-0.9.30.1/linux-2.6.30.10/arch/mips/bcm47xx/prom.c
3. Second build (this is soon fast)
4. Upload the firmware with tftp
5. First boot is OK (64MB RAM - with sdram_init=0x0009 as default, maybe from pin 9 ground restore).
6. I change nvram variables:
nvram set sdram_init=0x0011
nvram set sdram_ncdl=0
nvram commit
7. type reboot, and openwrt recognize 128m memory.. I can use reboot command and it works well, but I can't unplug the power cable, it would just brick and need a pin 9 ground restore.

by the way, I also can't get atheros card work for me at this box, as I mentioned here.

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Hi!

I think in step 5 you have sdram_init=0x000b for 64 MB RAM. Thist is the value after PIN9, and after clear-nvram, recovery firmware upload too.
For cable unplug:
Have you changed the /etc/init.d/nvram?
By default the nvram script change the sdram_init to 0x0009 for WL-500gP.

vargalex

poisonhzkj wrote:

I my case, with wl500gp 128m ram mod.

1. Build a firmware from trunk
2. In build_dir manually change the build_dir/toolchain-mipsel_gcc-4.3.3+cs_uClibc-0.9.30.1/linux-2.6.30.10/arch/mips/bcm47xx/prom.c
3. Second build (this is soon fast)
4. Upload the firmware with tftp
5. First boot is OK (64MB RAM - with sdram_init=0x0009 as default, maybe from pin 9 ground restore).
6. I change nvram variables:
nvram set sdram_init=0x0011
nvram set sdram_ncdl=0
nvram commit
7. type reboot, and openwrt recognize 128m memory.. I can use reboot command and it works well, but I can't unplug the power cable, it would just brick and need a pin 9 ground restore.

by the way, I also can't get atheros card work for me at this box, as I mentioned here.

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Hi vargalex,

I've tested your firmware (still without serial console), it DO function well with 128M ram! Any extra salt? I suppose we share the  same patch recipe.


vargalex wrote:

Hi Yellowbug!

Please check the firmware in #45 (https://forum.openwrt.org/viewtopic.php?pid=102330#p102330) for WL-500gP!

vargalex

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

yellowbug wrote:

Hi vargalex,

I've tested your firmware (still without serial console), it DO function well with 128M ram! Any extra salt? I suppose we share the  same patch recipe.

is it on rt-n16 ?

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

it's on 500gp with 128m mode. it should be okay on rt-n16 also except some different device drivers option.

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

yellowbug wrote:

it's on 500gp with 128m mode. it should be okay on rt-n16 also except some different device drivers option.

what are those drivers?>

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

USB host for example

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

do you know if wireless will work ?

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Hi!

Good news in ticket: https://dev.openwrt.org/ticket/3177

The problem was the 128MB RAM patch. By me the

for (mem = (1 << 20); mem < (128 << 20); mem += (1 << 20)) {

was OK. But by others not.

The right version is (this is in RT-N16 source too):

for (mem = (1 << 20); mem < (128 << 20); mem <<= 1) {

vargalex

yellowbug wrote:

Hi vargalex,

I've tested your firmware (still without serial console), it DO function well with 128M ram! Any extra salt? I suppose we share the  same patch recipe.


vargalex wrote:

Hi Yellowbug!

Please check the firmware in #45 (https://forum.openwrt.org/viewtopic.php?pid=102330#p102330) for WL-500gP!

vargalex

64

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Unfortunately applying changes from the patches to prom.c in trunk to both :
   build_dir/toolchain-mipsel_gcc-4.3.3+cs_uClibc-0.9.30.1/linux-2.6.32.9/arch/mips/bcm47xx/prom.c
and
  build_dir/linux-brcm47xx/linux-2.6.32.9/arch/mips/bcm47xx/prom.c
still doesn't seem to enable the RT-N16 to boot.

As always, i die at

Entry at 0x80001000
Starting program at 0x80001000

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

I have not tested, but a fixed 128MB RAM detection added to tunk: https://dev.openwrt.org/changeset/20072

vargalex

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

@tmg, did you test the trunk version from vargalex already on rt-n16?

@all: The new beta, is it running on rt-n16?

@me: shame, not time for trying on my rt-n16

67 (edited by yellowbug 2010-03-14 17:18:27)

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Has Anybody tested this patch already?
FYI, At least this patch did not function properly on my 500gp with 128m mod and led to a new brick(all led kept being off). Pin9/pin16 trick can force the router into diag mode, but the router keep bricked(all led kept being off) after any newly flashing, no mater which firmware were used. I'm still struggle to escape from the brick presently, any suggestion/idea will be appreciated.
I can't give futher more details by now due to the lack of serial console. I don't belive RT-N16 users would be more lucky then me, so be careful if u wanna try the trunk version.
Obviously, this problem could not be trigered with "sdram_init=0x0009".

vargalex wrote:

I have not tested, but a fixed 128MB RAM detection added to tunk: https://dev.openwrt.org/changeset/20072

vargalex

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Hi!


As I wrote, I have not tested. But first check your power supply too.

vargalex

yellowbug wrote:

Has Anybody tested this patch already?
FYI, At least this patch did not function properly on my 500gp with 128m mod and led to a new brick(all led kept being off). Pin9/pin16 trick can force the router into diag mode, but the router keep bricked(all led kept being off) after any newly flashing, no mater which firmware were used. I'm still struggle to escape from the brick presently, any suggestion/idea will be appreciated.
I can't give futher more details by now due to the lack of serial console. I don't belive RT-N16 users would be more lucky then me, so be careful if u wanna try the trunk version.
Obviously, this problem could not be trigered with "sdram_init=0x0009".

vargalex wrote:

I have not tested, but a fixed 128MB RAM detection added to tunk: https://dev.openwrt.org/changeset/20072

vargalex

69 (edited by yellowbug 2010-03-15 02:23:46)

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Power supply is definitely ok, at least they were ok when i did the pin-ground trick. sorry for missing some details in the privious post, i bricked two router. Both are the same symptom.

Edit: Lan/Wan led function well if connected, but power led keeps off.

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Hi!

Please write to ticket too: https://dev.openwrt.org/ticket/3177

Sorry for 2 routers, I hope you can repair them.

vargalex

yellowbug wrote:

Power supply is definitely ok, at least they were ok when i did the pin-ground trick. sorry for missing some details in the privious post, i bricked two router. Both are the same symptom.

Edit: Lan/Wan led function well if connected, but power led keeps off.

71 (edited by yellowbug 2010-03-16 17:00:09)

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Don't say sorry pls. It's my pleasure to test the cutting-edge submits for the community. Now I've unlocked the routers already -- For some reason,The CFE took tens of miniutes to finished flashing progress of bleeding edge snapshort, on the contrast, it took only miniutes on earlier version (backfire-beta for example) despite the slight fw size defference. So I had to flash backfire-beta first then use mtd to flash the snapshot, which took only miniutes.

Wish my information be useful to the developer team and comming explorers.

Regards


vargalex wrote:

Hi!

Please write to ticket too: https://dev.openwrt.org/ticket/3177

Sorry for 2 routers, I hope you can repair them.

vargalex

72 (edited by tmg 2010-03-19 01:56:37)

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

svn revision 20183 with the new patchs still does not boot for me on RT-N16. I have made my best guess at other CONFIG options, but still fail to even load the kernel.

Entry at 0x80001000
Starting program at 0x80001000

My hunch is that it has something to do with the trx header, . See http://www.bitsum.com/firmware_mod_kit.htm and http://svn.dd-wrt.com:8000/dd-wrt/browser/opt/asus . the bitsum version is later, but still only 2006.

I am looking particularly at the changesets from ddwrt http://svn.dd-wrt.com:8000/dd-wrt/changeset/13296  and http://svn.dd-wrt.com:8000/dd-wrt/changeset/13295 .

At some point I will try to manually build a trx file with a similar versioning strategy, but it may be a week, an I am not expert in such art. So be my guest if you wish.

All the best

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Hi!

I have readed by the ticket, that you forget to change the /etc/init.d/nvram script. But after changing everything is OK. Am I right?
In https://forum.openwrt.org/viewtopic.php?pid=102270#p102270 I wrote from this script error.


yellowbug wrote:

Don't say sorry pls. It's my pleasure to test the cutting-edge submits for the community. Now I've unlocked the routers already -- For some reason,The CFE took tens of miniutes to finished flashing progress of bleeding edge snapshort, on the contrast, it took only miniutes on earlier version (backfire-beta for example) despite the slight fw size defference. So I had to flash backfire-beta first then use mtd to flash the snapshot, which took only miniutes.

Wish my information be useful to the developer team and comming explorers.

Regards


vargalex wrote:

Hi!

Please write to ticket too: https://dev.openwrt.org/ticket/3177

Sorry for 2 routers, I hope you can repair them.

vargalex

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

Sorry, I really missed that. 500gp is ok presently. Let's wait for the further reports from RT-N16 users now.

vargalex wrote:

Hi!

I have readed by the ticket, that you forget to change the /etc/init.d/nvram script. But after changing everything is OK. Am I right?
In https://forum.openwrt.org/viewtopic.php?pid=102270#p102270 I wrote from this script error.

Re: Anyone heard of the Asus RT-N16? I think I have a rocker! BCM4718 128M

tmg wrote:

svn revision 20183 with the new patchs still does not boot for me on RT-N16. I have made my best guess at other CONFIG options, but still fail to even load the kernel.

Entry at 0x80001000
Starting program at 0x80001000

My hunch is that it has something to do with the trx header, . See http://www.bitsum.com/firmware_mod_kit.htm and http://svn.dd-wrt.com:8000/dd-wrt/browser/opt/asus . the bitsum version is later, but still only 2006.

I am looking particularly at the changesets from ddwrt http://svn.dd-wrt.com:8000/dd-wrt/changeset/13296  and http://svn.dd-wrt.com:8000/dd-wrt/changeset/13295 .

At some point I will try to manually build a trx file with a similar versioning strategy, but it may be a week, an I am not expert in such art. So be my guest if you wish.

All the best

Hi,

please see ticket 6580, where gregd72002 states:

gregd72002 wrote:

Some testing (conducted by 2 independent testers) pointed out that the reason for kernel dying while initialising is ssb initialisation (ssb scanning to be precise). This is just a suggestion as further testing has to be done.

Please note that this bug has nothing to do with TRX header and asus ADDVER. The bootloader loads the kernel without any fancy headers. Kernel gets initialised by discovering CPU, memory etc.. but ssb scanning seems to fail.