From: Grahame J. <gb...@th...> - 2010-08-27 07:07:49
Attachments:
TruFlo.jpg
|
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <meta http-equiv="content-type" content="text/html; charset=ISO-8859-1"> </head> <body bgcolor="#ffffff" text="#000000"> Hi,<br> <br> I have ported the Verdex to 2.6.34 because the BT layer has problems in 2.6.31 hoping to fix the issue.<br> <br> The kernel compiles and is configured correctly as far as I can see.<br> When the kernel boots from uboot the console disappears never to return.<br> I can ssh into the image and it runs my setup OK accept there are not serial terminals.<br> <br> Reality check:<br> .config as compile into kernel<br> <br> <tt>CONFIG_CMDLINE="console=ttyS0,115200n8 root=1f01 rootfstype=jffs2 debug"</tt><br> <tt><br> #</tt><br> <tt># Non-8250 serial port support</tt><br> <tt>#</tt><br> <tt>CONFIG_SERIAL_PXA=y</tt><br> <tt>CONFIG_SERIAL_PXA_CONSOLE=y</tt><br> <tt>CONFIG_SERIAL_CORE=y</tt><br> <tt>CONFIG_SERIAL_CORE_CONSOLE=y</tt><br> <tt>CONFIG_CONSOLE_POLL=y</tt><br> <tt># CONFIG_SERIAL_TIMBERDALE is not set</tt><br> <tt>CONFIG_UNIX98_PTYS=y</tt><br> <br> <br> <br> Boot from serial consoel with uboot<br> <br> <tt>GUM> katload 100000<br> Copying kernel to 0xa2000000 from 0x00f00000 (length 0x00100000)...done<br> GUM> bootm<br> ## Booting image at a2000000 ...<br> Image Name: Angstrom/2.6.34/gumstix-verdex<br> Image Type: ARM Linux Kernel Image (uncompressed)<br> Data Size: 1012652 Bytes = 988.9 kB<br> Load Address: a0008000<br> Entry Point: a0008000<br> OK<br> <br> Starting kernel ...<br> <br> Uncompressing Linux... done, booting the kernel.</tt><br> <br> That's it, no more output<br> <br> No /dev/ttyS?<br> ssh root@gumstix<br> <br> <tt>root@gumstix:~# ls -la /dev</tt><br> <tt>drwxr-xr-x 7 root root 660 Jan 1 10:04 .</tt><br> <tt>drwxr-xr-x 16 root root 0 Jan 1 1970 ..</tt><br> <tt>drwxr-xr-x 5 root root 120 Jan 1 10:04 .udev</tt><br> <tt>drwxr-xr-x 2 root root 480 Jan 1 10:04 char</tt><br> <tt>crw------- 1 root root 5, 1 Jan 1 10:05 console</tt><br> <tt>crw-rw---- 1 root root 10, 63 Jan 1 10:05 cpu_dma_latency</tt><br> <tt>crw-rw-rw- 1 root root 1, 7 Jan 1 10:05 full</tt><br> <tt>crw-rw---- 1 root root 89, 0 Jan 1 10:04 i2c-0</tt><br> <tt>crw-rw---- 1 root root 89, 1 Jan 1 10:04 i2c-1</tt><br> <tt>prw------- 1 root root 0 Jan 1 10:04 initctl</tt><br> <tt>crw-rw---- 1 root root 1, 11 Jan 1 10:05 kmsg</tt><br> <tt>srw-rw-rw- 1 root root 0 Jan 1 10:04 log</tt><br> <tt>crw-r----- 1 root kmem 1, 1 Jan 1 10:05 mem</tt><br> <tt>drwxr-xr-x 2 root root 60 Jan 1 10:04 misc</tt><br> <tt>crw-rw---- 1 root root 90, 0 Jan 1 10:04 mtd0</tt><br> <tt>crw-rw---- 1 root root 90, 1 Jan 1 10:04 mtd0ro</tt><br> <tt>crw-rw---- 1 root root 90, 2 Jan 1 10:04 mtd1</tt><br> <tt>crw-rw---- 1 root root 90, 3 Jan 1 10:04 mtd1ro</tt><br> <tt>crw-rw---- 1 root root 90, 4 Jan 1 10:04 mtd2</tt><br> <tt>crw-rw---- 1 root root 90, 5 Jan 1 10:04 mtd2ro</tt><br> <tt>crw-rw---- 1 root root 10, 62 Jan 1 10:05 network_latency</tt><br> <tt>crw-rw---- 1 root root 10, 61 Jan 1 10:05 network_throughput</tt><br> <tt>crw-rw-rw- 1 root root 1, 3 Jan 1 10:05 null</tt><br> <tt>crw-rw-rw- 1 root root 5, 2 Jan 1 10:40 ptmx</tt><br> <tt>drwxr-xr-x 2 root root 0 Jan 1 1970 pts</tt><br> <tt>crw-rw-rw- 1 root root 1, 8 Jan 1 10:05 random</tt><br> <tt>lrwxrwxrwx 1 root root 4 Jan 1 10:04 rtc -> rtc0</tt><br> <tt>crw-rw---- 1 root audio 254, 0 Jan 1 10:04 rtc0</tt><br> <tt>drwxrwxrwt 2 root root 40 Jan 1 10:05 shm</tt><br> <tt>crw-rw-rw- 1 root root 5, 0 Jan 1 10:06 tty</tt><br> <tt>-rw-r--r-- 1 root root 11 Jan 1 10:04 udev_network_queue</tt><br> <tt>crw-rw-rw- 1 root root 1, 9 Jan 1 10:05 urandom</tt><br> <tt>crw-rw-rw- 1 root root 1, 5 Jan 1 10:05 zero</tt><br> <br> <br> <tt>root@gumstix:~# cat /proc/cmdline <br> console=ttyS0,115200n8 root=1f01 rootfstype=jffs2 reboot=cold,hard<br> </tt><br> What has happened to the serial console?<br> <br> Thanks<br> <br> Grahame Jordan<br> <br> <div class="moz-signature">-- <br> <title></title> <font color="#00559f"><font face="Helvetica, sans-serif"><font style="font-size: 11pt;" size="2"><b>Glass Expansion</b></font></font></font><br> <font color="#000000" face="Helvetica, sans-serif"><font style="font-size: 10pt;" size="2">15 Batman Street</font></font><br> <font color="#000000" face="Helvetica, sans-serif"><font style="font-size: 10pt;" size="2">West Melbourne</font></font><br> <font color="#000000" face="Helvetica, sans-serif"><font style="font-size: 10pt;" size="2">Victoria 3003</font></font><br> <font color="#000000" face="Helvetica, sans-serif"><font style="font-size: 10pt;" size="2">AUSTRALIA</font></font><br> <font color="#000000" face="Helvetica, sans-serif"><font style="font-size: 10pt;" size="2">T: +61 3 9320 1111</font></font><br> <font color="#000000" face="Helvetica, sans-serif"><font style="font-size: 10pt;" size="2">F: +61 3 9320 1112</font></font><br> <font color="#00559f"><u><a href="http://www.geicp.com/"><font color="#00559f"><font face="Helvetica, sans-serif"><font style="font-size: 10pt;" size="2">www.geicp.com</font></font></font></a></u></font> <p> <a href="http://www.geicp.com/cgi-bin/site/wrapper.pl?c1=Updates_newproducts_truflo_sample_monitor"><img src="cid:par...@th..." border="0" width="140"></a> <br> <a href="http://www.geicp.com/cgi-bin/site/wrapper.pl?c1=Updates_newproducts_truflo_sample_monitor"><font style="font-size: 11pt;" color="#ff6633" face="Helvetica, sans-serif" size="2"><b>Click here to check out our new TruFlo Sample Monitor</b></font></a> </p> </div> </body> </html> |
From: Ash C. <as...@gu...> - 2010-08-30 16:35:36
|
Hi Grahame, On Fri, Aug 27, 2010 at 12:07 AM, Grahame Jordan <gb...@th...>wrote: > I have ported the Verdex to 2.6.34 because the BT layer has problems in > 2.6.31 hoping to fix the issue. > Cool. Did 2.6.34 fix the problems? What were the issues? Would you be willing to send me the kernel so I can give it a spin? > > The kernel compiles and is configured correctly as far as I can see. > When the kernel boots from uboot the console disappears never to return. > I can ssh into the image and it runs my setup OK accept there are not > serial terminals. > > Reality check: > .config as compile into kernel > > CONFIG_CMDLINE="console=ttyS0,115200n8 root=1f01 rootfstype=jffs2 debug" > > # > # Non-8250 serial port support > # > CONFIG_SERIAL_PXA=y > CONFIG_SERIAL_PXA_CONSOLE=y > CONFIG_SERIAL_CORE=y > CONFIG_SERIAL_CORE_CONSOLE=y > CONFIG_CONSOLE_POLL=y > # CONFIG_SERIAL_TIMBERDALE is not set > CONFIG_UNIX98_PTYS=y > This looks fine to me (although I don't recognize the CONSOLE_POLL). I think the 8250 stuff should likewise be disabled or built as a module so as not to interfere. In your conf/machine/gumstix-verdex.conf (or whatever) file, how are these variables set? USE_VT & SYSVINIT_ENABLED_GETTYS -Ash |
From: Jose L. R. <jr...@ir...> - 2010-08-31 00:07:04
|
El 30/08/10 18:35, Ash Charles escribió: > Hi Grahame, > > On Fri, Aug 27, 2010 at 12:07 AM, Grahame Jordan<gb...@th...>wrote: > >> I have ported the Verdex to 2.6.34 because the BT layer has problems in >> 2.6.31 hoping to fix the issue. >> I was working on porting the 2.6.34 to the Verdex platform during the last days. In my case, trying to fix the problems with the wireless exposed in: http://old.nabble.com/-verdex--libertas-wifi-driver-hangs -td28842840.html > Cool. Did 2.6.34 fix the problems? What were the issues? Would you be > willing to send me the kernel so I can give it a spin? If it can help, I migrate the 2.6.31 verdex patchset. The recipe, patch and defconfig can be found at: - http://www.iri.upc.edu/people/jrivero/gumstix/gumstix-kernel_2.6.34.bb - http://www.iri.upc.edu/people/jrivero/gumstix/gumstix-kernel-2.6.34.tar.gz The download has only one big patch, but I have a git repo with 2.6.31 patches ported individually which should be used when the things work. >> >> The kernel compiles and is configured correctly as far as I can see. >> When the kernel boots from uboot the console disappears never to return. I have the same problem. >> I can ssh into the image and it runs my setup OK accept there are not >> serial terminals. >> >> Reality check: >> .config as compile into kernel >> >> CONFIG_CMDLINE="console=ttyS0,115200n8 root=1f01 rootfstype=jffs2 debug" >> >> # >> # Non-8250 serial port support >> # >> CONFIG_SERIAL_PXA=y >> CONFIG_SERIAL_PXA_CONSOLE=y >> CONFIG_SERIAL_CORE=y >> CONFIG_SERIAL_CORE_CONSOLE=y >> CONFIG_CONSOLE_POLL=y >> # CONFIG_SERIAL_TIMBERDALE is not set >> CONFIG_UNIX98_PTYS=y >> > This looks fine to me (although I don't recognize the CONSOLE_POLL). I > think the 8250 stuff should likewise be disabled or built as a module so as > not to interfere. > > In your conf/machine/gumstix-verdex.conf (or whatever) file, how are these > variables set? USE_VT& SYSVINIT_ENABLED_GETTYS > In my case: - USE_VT = "1" - SYSVINIT_ENABLED_GETTYS = "1" If I found something interesting in the next days, I will post it here. Thanks. -- Jose Luis Rivero <jr...@ir...> Humanoid Lab. http://apollo.upc.es/humanoide |
From: Grahame J. <gb...@th...> - 2010-12-07 11:24:21
|
Has anyone solved this issue? I am still having trouble with anything greater than linux-2.6.33 Cheers Grahame Jordan On 09/01/2010 07:01 PM, José Luis Rivero wrote: > On 08/31/2010 09:14 PM, Ash Charles wrote: >> Hi, >> >> This is my summary about 2.6.34 based on this thread: >> - has the potential to solve a bluetooth problem >> - need a fix to deal with console problems >> - Grahame has a version (of the kernel or of the rootfs?) that fits in flash >> - some users are seeing an issue with the libertas driver in the >> current version. This issue still exists in 2.6.34 and may in fact be >> a low-level timing issue. >> >> Is this accurate? > Yees, but my main reason to migrate to kernel 2.6.34 was the possibility > to give a try to the thin firmware and libertas_cf driver released by > the OLPC guys which seems to work for the overo looking at thread: > > http://old.nabble.com/Marvell-88W8686-thin-firmware-released-for-SDIO-mode-td29100249.html > > Were you thinking in the firmware when talking about 'low-level timing > issue'? > >> Once we get a fix for the console issue, I'll happily accept patches >> and a 2.6.34 recipe for Verdex for the git repo. I can also pull >> changes from the Overo branch at that time. > No problema about preparing the patches but we still need to debug the > console problem. Bisecting the kernel having to port the patchset every > time could not be the funniest thing in the world ... =) > > Thanks Ash. > >> -Ash >> >> On Mon, Aug 30, 2010 at 5:55 PM, Grahame Jordan<gb...@th...> wrote: >>> I did some more testing and the console fails on 2.6.33 but still works on 2.6.32-21. >>> I cannot tell if the problem is solved as the BT requires a serial port :) >>> >>> However the BT problem seemed to go away when I started using 2.6.32. >>> >>> I have a suspicion that something fundamental has been changed in the underlying serial code in the kernel crossing over to 2.6.33. There a a few patches to the serial core here. >>> >>> I do a bitbake -c menuconfig to set the kernel parameters and find that OE gets in the way sometimes setting things I don't want set. >>> I have managed to get it to fit on flash also. >> ------------------------------------------------------------------------------ >> This SF.net Dev2Dev email is sponsored by: >> >> Show off your parallel programming skills. >> Enter the Intel(R) Threading Challenge 2010. >> http://p.sf.net/sfu/intel-thread-sfd >> _______________________________________________ >> gumstix-users mailing list >> gum...@li... >> https://lists.sourceforge.net/lists/listinfo/gumstix-users > |
From: Tomas T. <tta...@ge...> - 2011-06-10 07:21:51
|
On Tue, 2010-12-07 at 22:24 +1100, Grahame Jordan wrote: > Has anyone solved this issue? > > I am still having trouble with anything greater than linux-2.6.33 [SOLVED] The three additional lines shown in the patch below were removed as of linux-2.6.33. If you require UART access on the Gumstix verdex-pro (for console, bluetooth, etc.), they need to be restored. gumstix-verdex-pro-restore-uart.patch: diff --git a/arch/arm/mach-pxa/pxa27x.c b/arch/arm/mach-pxa/pxa27x.c index 909756e..ea5a3fd 100644 --- a/arch/arm/mach-pxa/pxa27x.c +++ b/arch/arm/mach-pxa/pxa27x.c @@ -413,6 +413,9 @@ void __init pxa27x_set_i2c_power_info(struct i2c_pxa_platform_data *info) static struct platform_device *devices[] __initdata = { &pxa27x_device_udc, + &pxa_device_ffuart, + &pxa_device_btuart, + &pxa_device_stuart, &pxa_device_pmu, &pxa_device_i2s, &pxa_device_asoc_ssp1, So the question is, who turned out the lights? :-P Best regards, Tomas Targownik |