busybox utility

[naveen yadav]

Dear all,

I am checking busybox utility, Is there any other package available
similar to busybox which  we can use in embedded system .

Thanks

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: busybox utility

[Thomas Petazzoni]

Le Tue, 22 Nov 2011 14:05:14 +0530,
naveen yadav yad.nav...@gmail.com a écrit :

 I am checking busybox utility, Is there any other package available
 similar to busybox which  we can use in embedded system .

This question is not kernel-related, so it shouldn't be posted on
linux-kernel@ or kernelnewbies@.

As an alternative to Busybox, you might want to have a look at
Embutils http://www.fefe.de/embutils/ or probably the Toolbox thing
from Android. But none of them have the same amount of features that
Busybox has.

Thomas
-- 
Thomas Petazzoni, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: LPC bus access howto

[Andre Haupt]

On Mon, Nov 21, 2011 at 05:53:30PM +0100, Guillaume Knispel wrote:
 On Mon, 21 Nov 2011 17:01:29 +0100
 Andre Haupt an...@bitwigglers.org wrote:
 
  Hi all,
  
  We have a FPGA which will be connected to the LPC bus. This FPGA will
  provide some GPIO functionality, a rotary encoder, etc.
  
  Now i want to write a kernel driver to talk to this FPGA.
  I am unsure how to get access to the LPC bus in the first place.
  What kernel APIs should i use? How to get the memory mapping right, etc?
 
 The LPC is typically automatically mapped by the hardware (southbridge)
 for some fixed ranges, various configurable ranges and might also use
 PCI substractive decoding. 
 
 LPC is indeed an ISA bus with the physical layer replaced by something
 with fewer signals, but from the point of view of the software there is
 almost no difference, so look for how to program ISA drivers and you
 will probably have your answer.

So request_mem_region(), ioremap(), ioread*() and iowrite*() are my
friends here right?

Still unsure how to obtain the LPC base address. Is this mapped to
0xA-0xF as with ISA?

cheers,

Andre

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: how set_user_nice() function is called

[Enes Albay - انس الباى]

Thanks for reply.

I also asked that set_user_nice where and when is called(file and line).  I
think it is not called after each forking operation, isn't it? Could you
give calling order between functions from do_fork() like that
func1()-func2()-...

Regards.

On Mon, Nov 21, 2011 at 3:59 AM, Mulyadi Santosa
mulyadi.sant...@gmail.comwrote:

 On Mon, Nov 21, 2011 at 00:31, Enes Albay - انس الباى
 albaye...@gmail.com wrote:
  Hi!
 
  anybody knows that where is set_user_nice() function called in the
 kernel?
  i couldn't find that.  i just found that in forking operation
  do_fork()-copy_process()-
  dup_task_struct() called then set_user_nice()
  called but i couldn't find that where. Also how/where do_fork() function
  called in kernel i couldn't find.

 In x86, do_fork() is called from sys_fork()... check :
 http://lxr.linux.no/#linux+v3.1.1/arch/x86/kernel/process.c#L235

 While sys_fork() itself, since it's a syscall handler, is called from
 entry.S. Please refer to:
 http://lxr.linux.no/#linux+v3.1.1/arch/x86/ia32/ia32entry.S#L475



 --
 regards,

 Mulyadi Santosa
 Freelance Linux trainer and consultant

 blog: the-hydra.blogspot.com
 training: mulyaditraining.blogspot.com




-- 
Enes Albay
___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

x86 kernel build chain - how bzImage gets created

[Graeme Russ]

Hi All,

Hopefully this will help demystify the build process

Notation:
primary input file (file size in bytes)
  |
  + additional input file
  |
  * command
  |
  \- output file (file size in bytes)

The file sizes are arbitrary and are based on the current configuration
options I have set for a small kernel. They are simply there to illustrate
the relative differences in file sizes.

Note that vmlinux.o is not used to generate bzImage but I've included it
for completeness

+ arch/x86/kernel/head_32.o
+ arch/x86/kernel/head32.o
+ arch/x86/kernel/head.o
+ arch/x86/kernel/init_task.o
+ init/built-in.o
+ usr/built-in.o
+ arch/x86/built-in.o
+ kernel/built-in.o
+ mm/built-in.o
+ fs/built-in.o
+ ipc/built-in.o
+ security/built-in.o
+ crypto/built-in.o
+ block/built-in.o
+ lib/lib.a
+ arch/x86/lib/lib.a
+ lib/built-in.o
+ arch/x86/lib/built-in.o
+ drivers/built-in.o
+ sound/built-in.o
+ firmware/built-in.o
+ arch/x86/pci/built-in.o
+ net/built-in.o
|
* ld -m elf_i386 -r -o vmlinux.o
|
\-vmlinux.o (2,889,050)

+ arch/x86/kernel/head_32.o
+ arch/x86/kernel/head32.o
+ arch/x86/kernel/head.o
+ arch/x86/kernel/init_task.o
+ init/built-in.o
+ usr/built-in.o
+ arch/x86/built-in.o
+ kernel/built-in.o
+ mm/built-in.o
+ fs/built-in.o
+ ipc/built-in.o
+ security/built-in.o
+ crypto/built-in.o
+ block/built-in.o
+ lib/lib.a
+ arch/x86/lib/lib.a
+ lib/built-in.o
+ arch/x86/lib/built-in.o
+ drivers/built-in.o
+ sound/built-in.o
+ firmware/built-in.o
+ arch/x86/pci/built-in.o
+ net/built-in.o
+ .tmp_kallsyms2.o
|
* ld -m elf_i386 --build-id -T arch/x86/kernel/vmlinux.lds
|
\-vmlinux (2,629,659)
   |
   * objcopy -R .comment -S
   |
   \-arch/x86/boot/compressed/vmlinux.bin (2,094,132)
  |
  * gzip
  |
  \- arch/x86/boot/compressed/vmlinux.bin.gz (1,074,711)
 |
 * .incbin
 |
 \- piggy.S
|
* gcc
|
\- piggy.o (1,075,419)
   |
   + arch/x86/boot/compressed/head_32.o
   + arch/x86/boot/compressed/misc.o
   + arch/x86/boot/compressed/string.o
   + arch/x86/boot/compressed/cmdline.o
   + arch/x86/boot/compressed/early_serial_console.o
   |
   * ld -m elf_i386 -T arch/x86/boot/compressed/vmlinux.lds
   |
   \-arch/x86/boot/compressed/vmlinux (1,099,538)
  |
  * objcopy -O binary -R .note -R .comment -S
  |
  \- arch/x86/boot/vmlinux.bin (1,092,060)
 |
 |   + arch/x86/boot/a20.o
 |   + arch/x86/boot/bioscall.o
 |   + arch/x86/boot/cmdline.o
 |   + arch/x86/boot/copy.o
 |   + arch/x86/boot/cpu.o
 |   + arch/x86/boot/cpucheck.o
 |   + arch/x86/boot/early_serial_console.o
 |   + arch/x86/boot/edd.o
 |   + arch/x86/boot/header.o
 |   + arch/x86/boot/main.o
 |   + arch/x86/boot/mca.o
 |   + arch/x86/boot/memory.o
 |   + arch/x86/boot/pm.o
 |   + arch/x86/boot/pmjump.o
 |   + arch/x86/boot/printf.o
 |   + arch/x86/boot/regs.o
 |   + arch/x86/boot/string.o
 |   + arch/x86/boot/tty.o
 |   + arch/x86/boot/video.o
 |   + arch/x86/boot/video-mode.o
 |   + arch/x86/boot/version.o
 |   + arch/x86/boot/video-vga.o
 |   + arch/x86/boot/video-vesa.o
 |   + arch/x86/boot/video-bios.o
 |   |
 |   * ld -m elf_i386 -T arch/x86/boot/setup.ld
 |   |
 |   \-arch/x86/boot/setup.elf (186,171)
 |  |
 |  * objcopy -O binary arch/x86/boot/setup.elf
 |  |
 +  \- arch/x86/boot/setup.bin (12,604)
 |
 * arch/x86/boot/tools/build
 |
 \- arch/x86/boot/bzImage (1,104,864)

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: x86 kernel build chain - how bzImage gets created

[Daniel Baluta]

On Tue, Nov 22, 2011 at 12:46 PM, Graeme Russ graeme.r...@gmail.com wrote:
 Hi All,

 Hopefully this will help demystify the build process

 Notation:
 primary input file (file size in bytes)
  |
  + additional input file
  |
  * command
  |
  \- output file (file size in bytes)

 The file sizes are arbitrary and are based on the current configuration
 options I have set for a small kernel. They are simply there to illustrate
 the relative differences in file sizes.

Great! Thanks for sharing this.

 Note that vmlinux.o is not used to generate bzImage but I've included it
 for completeness

[...]
 + net/built-in.o
 |
 * ld -m elf_i386 -r -o vmlinux.o
 |
 \-vmlinux.o (2,889,050)

What is the purpose of vmlinux.o then?

thanks,
Daniel.

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: x86: Executing a raw vmlinux image (embedded environment)

[Graeme Russ]

Thanks syed,

Ok, I did a little more digging...

On 22/11/11 16:34, sk.syed2 wrote:
   /vmlinux2,629,659  bytes
   /vmlinux.o  2,889,050  bytes
   /arch/i386/boot/bzImage 1,104,864  bytes
   /arch/x86/boot/bzImage  1,104,864  bytes
   /arch/x86/boot/vmlinux.bin  1,092,060  bytes
   /arch/x86/boot/compressed/vmlinux   1,099,538  bytes
   /arch/x86/boot/compressed/vmlinux.bin   2,094,132  bytes
   /arch/x86/boot/compressed/vmlinux.bin.gz1,074,711  bytes

 I understand that /arch/x86/boot/compressed/vmlinux.bin.gz is a compressed
 version of /arch/x86/boot/compressed/vmlinux.bin, and
 /arch/i386/boot/bzImage and /arch/x86/boot/bzImage are the same file and
 that it is the 16-bit boot code + /arch/x86/boot/compressed/vmlinux.bin.gz
 This is correct.
 but I don't understand the rest...

I traced it all out. I sent a separate message to the ML documenting the
bzImage build chain - It's rather fascinating

 

 My guess is that /vmlinux.o is the ELF image generated by the compiler +
 linker stage and /vmlinux may be /vmlinux.o objdump'd into a raw binary and
 perhaps /arch/x86/boot/vmlinux.bin is a further stripped version
 of/vmlinux, but I'm at a loss with /arch/x86/boot/compressed/vmlinux
 
 vmlinux is ELF image with ELF header. So actual point of kernel entry
 would be at an offset, somewhere after the ELF header.
 vmlinux.bin is what you would get after doing
 #objcopy -O binary vmlinux vmlinux.bin.
 vmlinux.bin has only obj code and nothing else.

To be more precise:

arch/x86/boot/compressed/vmlinux.bin is the result of
#objcopy -R .comment -S vmlinux
so it is still an ELF image

arch/x86/boot/vmlinux.bin is the result of:
#objcopy -O binary -R .note -R .comment -S arch/x86/boot/compressed/vmlinux

arch/x86/boot/compressed/vmlinux has the decompression stub + compressed
version of arch/x86/boot/compressed/vmlinux.bin


 In any event, it looks like either /arch/x86/boot/compressed/vmlinux.bin or
 /vmlinux is what I need to copy into RAM @ 0x10 (1MiB) which is where
 my non-relocatable kernel is compiled to.
 
 copy vmlinux.bin.

I don't think either is what I want - arch/x86/boot/vmlinux.bin still
contains a compressed kernel. And the contents of the compressed section is
an ELF image which requires more memcpys and memsets

What I really want, I think, is:

#objcopy -O binary -R .comment -S vmlinux

Which is not generated during the build process

  - How to setup the memory map (keeping in mind I have 2GB of contiguous
   memory with no BIOS/ACPI etc to worry about clobbering
  - Any other tricks I need to be aware of..
 Check if x86 kernel expects some parameters(like machineid, bootargs
 location etc) in some registers.
 Check if x86 has low level debug support(like DEBUG_LL).
 Also you might want to check how initial page tables are being setup in 
 kernel.

I need to have a good look at what goes into arch/x86/boot/setup.bin - That
will really tell me what I need to do

Thanks,

Graeme

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: x86 kernel build chain - how bzImage gets created

[Graeme Russ]

Hi Daniel

On 22/11/11 21:59, Daniel Baluta wrote:
 On Tue, Nov 22, 2011 at 12:46 PM, Graeme Russ graeme.r...@gmail.com wrote:
 Hi All,

 Hopefully this will help demystify the build process

 Notation:
 primary input file (file size in bytes)
  |
  + additional input file
  |
  * command
  |
  \- output file (file size in bytes)

 The file sizes are arbitrary and are based on the current configuration
 options I have set for a small kernel. They are simply there to illustrate
 the relative differences in file sizes.
 
 Great! Thanks for sharing this.

Your welcome :)

 
 Note that vmlinux.o is not used to generate bzImage but I've included it
 for completeness
 
 [...]
 + net/built-in.o
 |
 * ld -m elf_i386 -r -o vmlinux.o
 |
 \-vmlinux.o (2,889,050)
 
 What is the purpose of vmlinux.o then?

I don't know - Anyone?

Regards,

Graeme

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

RE: a few questions regarding acpi and kernel configuration

[Pritam Bankar]

Hi, 
To cut down drivers you should configure your .config file in your source.
You will have to recompile kernel.
For pata you have ,

CONFIG_PATA_AMD=y
CONFIG_PATA_OLDPIIX=y
CONFIG_PATA_SCH=y

This can be set to n instead of y to remover support of pata. This can
be done to other modules also.





-Original Message-
From: kernelnewbies-boun...@kernelnewbies.org
[mailto:kernelnewbies-boun...@kernelnewbies.org] On Behalf Of Littlefield,
Tyler
Sent: 19 October 2011 03:43
To: kernelnewbies@kernelnewbies.org
Subject: a few questions regarding acpi and kernel configuration

Hello all:
I have a few questions, I was hoping someone could help me out here.
First, I'll start off by saying that my netbook is a sort of PDA for me. 
I just boot it up and when I'm taking care of stuff for work running around
getting info and that, I can use it to grab the info. I use Linux for this,
but my battery dies kind of quickly. I'm also looking to optomize
performance as much as possible.
So, I sat down and tried to optomize a kernel. Both to save space (since
this is a smaller drive), and to make things a bit quicker, even by a little
bit.
I put the config here:
http://tds-solutions.net/config.tar.bz2
First, I'm curious what else I can yank out of there to make everything
smaller. I'd like to cut down on the drivers and all that, and just include
the driver for my specific harddrive if that's possible, as well as the
wireless card. Can I turn off scsi/pata/ata since I have a sata drive? Is
there other stuff I missed that I could take out?
Finally, my last two questions.
1) Is there a way to access acpi so I can put the computer into standby
quickly? I'm not sure how to pull that off from the cli, but doing it with
code would work.
2) Is there a way for me to detect when the lid of my laptop has been closed
and put it into standby, and/or set an idle timer to put it into standby? If
there are clients out there that can interface with acpi and help with this,
awesome. If not, and I need to code something up, that works too.
Finally, I'm trying to find a way, since I do not use it to turn off the
display. Is this possible?

-- 

Take care,
Ty
Web: http://tds-solutions.net
The Aspen project: a light-weight barebones mud engine
http://code.google.com/p/aspenmud

Sent from my toaster.


___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies


___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: Path for STOLEN Packets!!!

[shivanth]

On 11/22/2011 1:26 PM, Kesava Srinivas wrote:

Hi ,
Even after lot of google; didn't get the accurate answer for this. 
After STEALING Packets (NF_STOLEN); want to send the Manipulated 
Packet (added extra header) to some of the available net devices. 
Please let me know what is the Function that I Should call to do this 
step ?? Also; once the skb was expanded by 20 bytes to add new ip 
header; do I need to reassign the skb pointer to the new location ??


-Thanks in Advance,
VKS



___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Hi srinivas,
You can send/transmit  the manipulated skb buffer by passing it to 
dev_queue_xmit . Be carefull though, any error crafting the packet, a 
kernel panic is waiting 4 you :-)

These articles might help
http://www.phrack.org/issues.html?issue=61id=13 
http://www.phrack.org/issues.html?issue=61id=13


http://www.phrack.org/issues.html?issue=55id=12#article
Regards
shivanth

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re:

[Jimmy Pan]

Sorry for the spam here, my email account was stolen.


On Mon, Nov 21, 2011 at 11:22 PM, Jimmy Pan dsp...@gmail.com wrote:
 ..Do you want to feel something new? Do you want to feel new
 unforgettable sensations? This is for you!
 http://un-ocean.fr/p.g.php?wellink_friend_id=14ox0


___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: x86 kernel build chain - how bzImage gets created

[Mulyadi Santosa]

On Tue, Nov 22, 2011 at 17:46, Graeme Russ graeme.r...@gmail.com wrote:
 Hi All,

 Hopefully this will help demystify the build process


Hi!

Wow great man! I deeply suggest to write that into kernelnewbies wiki too

Once again, great job!

-- 
regards,

Mulyadi Santosa
Freelance Linux trainer and consultant

blog: the-hydra.blogspot.com
training: mulyaditraining.blogspot.com

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Serial console temporarily garbled

[Neil Baylis]

I'm working on an embedded system running a 2.6.32 kernel originally
derived from RHEL6.

The hardware architecture is similar to a PC. The device is a specialized
server. It uses an 8250 compatible serial port for the console.

When it's booting up, everything looks fine until it initializes the serial
driver. At that point the output to the console becomes garbled until (I
think...) getty runs for the console port. Then it's readable again.

It looks as though the initialization of the serial driver is setting the
wrong baud rate, or perhaps some other configuration. During the boot
process, this garbled output happens for about 1 second, during which
probably a few thousand characters of garbage are output.

The grub boot parameter specifies console=ttys0,57600
The getty line for the console specifies T0:23:respawn:/sbin/getty -L
ttys0 57600 vt100

It's not clear to me how the serial driver is supposed to be initialized to
the right baud rate. Is the parameter from the boot line supposed to be
used?

(Just to complicate matters, we're using a Debian userspace with this
kernel).

Here's how it looks on the console:

[4.429923] Non-volatile memory driver v1.3

[4.438325] Serial: 8250/16550 driver, 4 ports, IRQ sharing enabled

ÿ[4.615305] serial8250: ttyS0 at I/O 0x3f8 (irq = 4) is a ST16650V2

;s¶6T=$É­½¤©;s¶v´=ĽÁÔP%1D²äÍÓÉK æÑ- %1¹©;s¶5V=ÔÍê£!¬E¹)¥°?

U000ÉP%19)$%°©;s61v=¡ 0¦0¦1r2CÔJ °æÁ0424¬Ñ6/F3d °0 ÕÁ4!



Much garbage deleted...



ؤµ!V°ÉD¨«Í'µõ;Ðs°W¶G=dÉ©¥®5ÕV$´Ô¥½Y§ ³°V%jê;Ðs1·Ë=t²E%²ÑiÑ)¤D!6f¸ÕêLoading,
please wait...

PuTTYPuTTYPuTTYBegin: Loading essential drivers ... done.

Begin: Running /scripts/init-premount ...
PuTTYPuTTYPuTTYPuTTYPuTTYPuTTYPuTTYPuTTYdone.

Begin: Mounting root file system ... Begin: Running /scripts/local-top ...
done.

If anyone can point me in the right direction to understand this, that
would be great. I think the repeated 'PuTTY' strings result from control
characters being sent to the terminal, which is PuTTY. It's answerback
string is 'PuTTY' by default.

Thanks,

Neil Baylis
___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: Serial console temporarily garbled

[Dave Hylands]

Hi Neil,

On Tue, Nov 22, 2011 at 10:20 AM, Neil Baylis neil.bay...@gmail.com wrote:
 I'm working on an embedded system running a 2.6.32 kernel originally derived
 from RHEL6.

 The hardware architecture is similar to a PC. The device is a specialized
 server. It uses an 8250 compatible serial port for the console.

 When it's booting up, everything looks fine until it initializes the serial
 driver. At that point the output to the console becomes garbled until (I
 think...) getty runs for the console port. Then it's readable again.

 It looks as though the initialization of the serial driver is setting the
 wrong baud rate, or perhaps some other configuration. During the boot
 process, this garbled output happens for about 1 second, during which
 probably a few thousand characters of garbage are output.

 The grub boot parameter specifies console=ttys0,57600

Did you mean ttyS0? The capital S is significant.

The serial ports default to 9600 baud, which is probably what the
garbage coming out is.

 The getty line for the console specifies T0:23:respawn:/sbin/getty -L ttys0
 57600 vt100

This should probably also be ttyS0

-- 
Dave Hylands
Shuswap, BC, Canada
http://www.davehylands.com

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: Serial console temporarily garbled

[Neil Baylis]

My apologies for the confusion, I meant ttyS0. I mistyped when composing
the email.

Neil

On Tue, Nov 22, 2011 at 10:43 AM, Dave Hylands dhyla...@gmail.com wrote:

 Hi Neil,

 On Tue, Nov 22, 2011 at 10:20 AM, Neil Baylis neil.bay...@gmail.com
 wrote:
  I'm working on an embedded system running a 2.6.32 kernel originally
 derived
  from RHEL6.
 
  The hardware architecture is similar to a PC. The device is a specialized
  server. It uses an 8250 compatible serial port for the console.
 
  When it's booting up, everything looks fine until it initializes the
 serial
  driver. At that point the output to the console becomes garbled until (I
  think...) getty runs for the console port. Then it's readable again.
 
  It looks as though the initialization of the serial driver is setting the
  wrong baud rate, or perhaps some other configuration. During the boot
  process, this garbled output happens for about 1 second, during which
  probably a few thousand characters of garbage are output.
 
  The grub boot parameter specifies console=ttys0,57600

 Did you mean ttyS0? The capital S is significant.

 The serial ports default to 9600 baud, which is probably what the
 garbage coming out is.

  The getty line for the console specifies T0:23:respawn:/sbin/getty -L
 ttys0
  57600 vt100

 This should probably also be ttyS0

 --
 Dave Hylands
 Shuswap, BC, Canada
 http://www.davehylands.com

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

GPL-only symbol Error

[Sengottuvelan S]

Hi All,

Basically, I am developing a kernel module named xyz.ko  and my
MODULE_LICENSE is Proprietary. I have got an error GPL-incompatible
module xyz.ko uses GPL-only symbols.

I am having trouble identifying it. How to avoid this error in during
MODPOST?. Is there any way to avoid this error?

Can you someone throw light on this.

Regards,
S. Sengottuvelan.
___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

RE: GPL-only symbol Error

[Jeff Haran]

 

 

From: kernelnewbies-boun...@kernelnewbies.org
[mailto:kernelnewbies-boun...@kernelnewbies.org] On Behalf Of
Sengottuvelan S
Sent: Tuesday, November 22, 2011 1:49 PM
To: Kernel Newbies
Subject: GPL-only symbol Error

 

Hi All,

 

Basically, I am developing a kernel module named xyz.ko  and my
MODULE_LICENSE is Proprietary. I have got an error GPL-incompatible
module xyz.ko uses GPL-only symbols.

 

I am having trouble identifying it. How to avoid this error in during
MODPOST?. Is there any way to avoid this error? 

 

Can you someone throw light on this.


Regards,
S. Sengottuvelan.

 

Legally, I believe your only option is to change your code to not
reference the EXPORT_SYMBOL_GPL () symbols that it is currently
referencing.

 

Jeff Haran

 


___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: GPL-only symbol Error

[Sengottuvelan S]

Hi Jeff,

I have few symbols with this error especially - sock_recv_timestamp(). Is
there way to refer this or similar API in my module with MODULE_LICENSE is
Proprietary. How do I know equavalent non-EXPORT_SYMBOL_GPL  APIs to
achieve similar functionality.
Regards
Sengottuvelan.S
On Tue, Nov 22, 2011 at 1:53 PM, Jeff Haran jha...@bytemobile.com wrote:

 ** **

 ** **

 *From:* kernelnewbies-boun...@kernelnewbies.org [mailto:
 kernelnewbies-boun...@kernelnewbies.org] *On Behalf Of *Sengottuvelan S
 *Sent:* Tuesday, November 22, 2011 1:49 PM
 *To:* Kernel Newbies
 *Subject:* GPL-only symbol Error

 ** **

 Hi All,

  

 Basically, I am developing a kernel module named xyz.ko  and my
 MODULE_LICENSE is Proprietary. I have got an error GPL-incompatible
 module xyz.ko uses GPL-only symbols.

  

 I am having trouble identifying it. How to avoid this error in during
 MODPOST?. Is there any way to avoid this error? 

  

 Can you someone throw light on this.


 Regards,
 S. Sengottuvelan.

 ** **

 Legally, I believe your only option is to change your code to not
 reference the EXPORT_SYMBOL_GPL () symbols that it is currently referencing.
 

 ** **

 Jeff Haran

 ** **




-- 
Regards,
S. Sengottuvelan.
___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

RE: GPL-only symbol Error

[Jeff Haran]

From: Sengottuvelan S [mailto:sengottuvela...@gmail.com] 
Sent: Tuesday, November 22, 2011 2:09 PM
To: Jeff Haran
Cc: Kernel Newbies
Subject: Re: GPL-only symbol Error

Hi Jeff,
 
I have few symbols with this error especially - sock_recv_timestamp(). Is there 
way to refer this or similar API in my module with MODULE_LICENSE is 
Proprietary. How do I know equavalent non-EXPORT_SYMBOL_GPL  APIs to achieve 
similar functionality.
Regards
Sengottuvelan.S
On Tue, Nov 22, 2011 at 1:53 PM, Jeff Haran jha...@bytemobile.com wrote:
 From: kernelnewbies-boun...@kernelnewbies.org 
[mailto:kernelnewbies-boun...@kernelnewbies.org] On Behalf Of Sengottuvelan S
Sent: Tuesday, November 22, 2011 1:49 PM
To: Kernel Newbies
Subject: GPL-only symbol Error
 Hi All,
 Basically, I am developing a kernel module named xyz.ko  and my 
MODULE_LICENSE is Proprietary. I have got an error GPL-incompatible module 
xyz.ko uses GPL-only symbols.
I am having trouble identifying it. How to avoid this error in during MODPOST?. 
Is there any way to avoid this error? 
Can you someone throw light on this.
Regards,
S. Sengottuvelan.
Legally, I believe your only option is to change your code to not reference the 
EXPORT_SYMBOL_GPL () symbols that it is currently referencing.
Jeff Haran

sock_recv_timestamp() is an inline function, but it calls 
__sock_recv_timestamp() which is declared like so:

EXPORT_SYMBOL_GPL(__sock_recv_timestamp);

And that is I suspect what is generating the error.

You can try to find equivalent functionality in other kernel code that is not 
so declared. Good luck with that. Even if you do find it, don't be surprised if 
down the line you decide to upgrade the kernel you are using and then you find 
that what were EXPORT_SYMBOL() symbols have been changed to EXPORT_SYMBOL_GPL() 
and you'll have to do it all over again, if it turns out to be even possible. 
That's a bad place to be in since its usually completely unanticipated work. 
Blows development schedules right out of the water.

I've seen others when faced with this who build their own kernels from sources 
just modify the problematic EXPORT_SYMBOL_GPL()s to EXPORT_SYMBOL()s. I don't 
know if that is legal. I wouldn't do it personally. Consult a lawyer before you 
go down that road.

This stuff is designed to encourage you and more specifically the company you 
work for to declare your module as GPL and make it available for others to use. 
That's how Linux grows. Convincing management to do so can be quite challenging 
though. If your company must keep its intellectual property private, they can 
always dump Linux and use one of the BSD derivatives instead that aren't 
covered by GPL.

Good luck,

Jeff Haran




___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: GPL-only symbol Error

[Greg KH]

On Tue, Nov 22, 2011 at 02:08:59PM -0800, Sengottuvelan S wrote:
 Hi Jeff,
  
 I have few symbols with this error especially - sock_recv_timestamp(). Is 
 there
 way to refer this or similar API in my module with MODULE_LICENSE is
 Proprietary. How do I know equavalent non-EXPORT_SYMBOL_GPL  APIs to achieve
 similar functionality.

There are none, you need to fix the license of your kernel module.

You do realize the kernel developer community's stance on non-GPL Linux
kernel modules, right?  Any reason why your company feels they can go
against their wishes?

thanks,

greg k-h

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: GPL-only symbol Error

[Greg KH]

On Tue, Nov 22, 2011 at 02:35:24PM -0800, Jeff Haran wrote:
 I've seen others when faced with this who build their own kernels from
 sources just modify the problematic EXPORT_SYMBOL_GPL()s to
 EXPORT_SYMBOL()s. I don't know if that is legal. I wouldn't do it
 personally. Consult a lawyer before you go down that road.

It is not legal and companies have gotten into big trouble by trying to
do that, or by creating gpl-condom kernel modules that wrap gpl-only
symbols and export them again.  Do not do that without the full buy-in
from your legal department as they do not want to hear about it from an
external query first.

 This stuff is designed to encourage you and more specifically the
 company you work for to declare your module as GPL and make it
 available for others to use. That's how Linux grows. Convincing
 management to do so can be quite challenging though. If your company
 must keep its intellectual property private, they can always dump
 Linux and use one of the BSD derivatives instead that aren't covered
 by GPL.

Exactly, if you want to use Linux, you must abide by the license of it,
just like any body of software.  What would happen if you decided to
ignore the license of Microsoft's operating system?  :)

thanks,

greg k-h

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

RE: GPL-only symbol Error

[Jeff Haran]

 -Original Message-
 From: Greg KH [mailto:g...@kroah.com]
 Sent: Tuesday, November 22, 2011 2:44 PM
 To: Jeff Haran; Sengottuvelan S; Kernel Newbies
 Subject: Re: GPL-only symbol Error
 
 On Tue, Nov 22, 2011 at 02:35:24PM -0800, Jeff Haran wrote:
  I've seen others when faced with this who build their own kernels
from
  sources just modify the problematic EXPORT_SYMBOL_GPL()s to
  EXPORT_SYMBOL()s. I don't know if that is legal. I wouldn't do it
  personally. Consult a lawyer before you go down that road.
 
 It is not legal and companies have gotten into big trouble by trying
to
 do that, or by creating gpl-condom kernel modules that wrap gpl-only
 symbols and export them again.  Do not do that without the full buy-in
 from your legal department as they do not want to hear about it from
an
 external query first.
 

Greg,

Just curious, can you provide links to these cases?

I've read the COPYING file at the top of the Linux source tree. I am not
a lawyer, but I don't see anything in it that would prohibit somebody
from taking the GPL kernel sources, changing the EXPORT_SYMBOL_GPL()s to
EXPORT_SYMBOL()s, publishing that modified kernel source as required by
the GPL license but then keep their module source that uses the now
non-GPL symbols private. It seems like it should be prohibited in the
spirit of open source, but I don't see any mention of these symbol
declarations in the license.

Thanks,

Jeff Haran




___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: GPL-only symbol Error

[Greg KH]

On Tue, Nov 22, 2011 at 04:34:21PM -0800, Jeff Haran wrote:
 Just curious, can you provide links to these cases?

No kernel cases ever became public, but the SAMBA team have lots of
public cases where they have successfully handled this type of case.

 I've read the COPYING file at the top of the Linux source tree. I am not
 a lawyer, but I don't see anything in it that would prohibit somebody
 from taking the GPL kernel sources, changing the EXPORT_SYMBOL_GPL()s to
 EXPORT_SYMBOL()s, publishing that modified kernel source as required by
 the GPL license but then keep their module source that uses the now
 non-GPL symbols private. It seems like it should be prohibited in the
 spirit of open source, but I don't see any mention of these symbol
 declarations in the license.

The license does not say anything about this, nor should it.

thanks,

greg k-h

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: GPL-only symbol Error

[Graeme Russ]

Hi Jeff,

On Wed, Nov 23, 2011 at 11:34 AM, Jeff Haran jha...@bytemobile.com wrote:
 -Original Message-
 From: Greg KH [mailto:g...@kroah.com]
 Sent: Tuesday, November 22, 2011 2:44 PM
 To: Jeff Haran; Sengottuvelan S; Kernel Newbies
 Subject: Re: GPL-only symbol Error

 On Tue, Nov 22, 2011 at 02:35:24PM -0800, Jeff Haran wrote:
  I've seen others when faced with this who build their own kernels
 from
  sources just modify the problematic EXPORT_SYMBOL_GPL()s to
  EXPORT_SYMBOL()s. I don't know if that is legal. I wouldn't do it
  personally. Consult a lawyer before you go down that road.

 It is not legal and companies have gotten into big trouble by trying
 to
 do that, or by creating gpl-condom kernel modules that wrap gpl-only
 symbols and export them again.  Do not do that without the full buy-in
 from your legal department as they do not want to hear about it from
 an
 external query first.


 Greg,

 Just curious, can you provide links to these cases?

 I've read the COPYING file at the top of the Linux source tree. I am not
 a lawyer, but I don't see anything in it that would prohibit somebody
 from taking the GPL kernel sources, changing the EXPORT_SYMBOL_GPL()s to
 EXPORT_SYMBOL()s, publishing that modified kernel source as required by
 the GPL license but then keep their module source that uses the now
 non-GPL symbols private. It seems like it should be prohibited in the
 spirit of open source, but I don't see any mention of these symbol
 declarations in the license.

The mere fact that the gpl-condom module links to EXPORT_SYMBOL_GPL
functions make it, in and of itself, a GPL module

ergo, there is no such thing as a gpl-condom module

Regards,

Graeme

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: GPL-only symbol Error

[Guillaume Knispel]

On Tue, 22 Nov 2011 17:21:46 -0800
Jeff Haran jha...@bytemobile.com wrote:

  -Original Message-
  From: Graeme Russ [mailto:graeme.r...@gmail.com]
  Sent: Tuesday, November 22, 2011 5:10 PM
  To: Jeff Haran
  Cc: Greg KH; Sengottuvelan S; Kernel Newbies
  Subject: Re: GPL-only symbol Error
  
  Hi Jeff,
  
  On Wed, Nov 23, 2011 at 11:34 AM, Jeff Haran jha...@bytemobile.com
  wrote:
   -Original Message-
   From: Greg KH [mailto:g...@kroah.com]
   Sent: Tuesday, November 22, 2011 2:44 PM
   To: Jeff Haran; Sengottuvelan S; Kernel Newbies
   Subject: Re: GPL-only symbol Error
  
   On Tue, Nov 22, 2011 at 02:35:24PM -0800, Jeff Haran wrote:
I've seen others when faced with this who build their own kernels
   from
sources just modify the problematic EXPORT_SYMBOL_GPL()s to
EXPORT_SYMBOL()s. I don't know if that is legal. I wouldn't do it
personally. Consult a lawyer before you go down that road.
  
   It is not legal and companies have gotten into big trouble by trying
   to
   do that, or by creating gpl-condom kernel modules that wrap gpl-only
   symbols and export them again.  Do not do that without the full buy-in
   from your legal department as they do not want to hear about it from
   an
   external query first.
  
  
   Greg,
  
   Just curious, can you provide links to these cases?
  
   I've read the COPYING file at the top of the Linux source tree. I am not
   a lawyer, but I don't see anything in it that would prohibit somebody
   from taking the GPL kernel sources, changing the EXPORT_SYMBOL_GPL()s
  to
   EXPORT_SYMBOL()s, publishing that modified kernel source as required by
   the GPL license but then keep their module source that uses the now
   non-GPL symbols private. It seems like it should be prohibited in the
   spirit of open source, but I don't see any mention of these symbol
   declarations in the license.
  
  The mere fact that the gpl-condom module links to EXPORT_SYMBOL_GPL
  functions make it, in and of itself, a GPL module
  
  ergo, there is no such thing as a gpl-condom module
  
  Regards,
  
  Graeme
 
 Graeme,
 
 Perhaps, but that's not what I asked about. It seems to me the essence of GPL 
 is that it grants people the right to modify GPL sources like the Linux 
 kernel in any way they want so long as they make those changes available to 
 whoever uses the code in the future. I don't see anything in it that 
 prohibits specific changes. So if I take a symbol that in the sources from 
 kernel.org is declared with EXPORT_SYMBOL_GPL(), make a 1 line change that 
 declares it EXPORT_SYMBOL() and put that on a publically available web site, 
 how have I violated GPL?
 
 Let's say I then ship a product that uses that custom kernel and a non-GPL 
 kernel module of my own writing that only works with the custom kernel, how 
 is that prohibited in the GPL license?
 
 Not that I am planning on doing this and I've never done it in the past, but 
 technically it seems that there would be no violation here.

I guess you can't really take some GPL code from a third party, do some
random paperwork or magic trick that basically constitute an unilateral
declaration from you that what was once considered and clearly
identified as derivatives are not anymore, publish said derivatives
that you pretend are not and do so in an incompatible licence, and get
along with it. On the contrary such behavior would very probably
constitute something like willfulness infringement if such thing is
applicable (but IANAL and so over...)

-- 
Guillaume Knispel
Avencall - 10 bis, rue Lucien Voilin - 92800 Puteaux
Tel. : (+33) 141 389 960

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: GPL-only symbol Error

[Greg Freemyer]

On Tue, Nov 22, 2011 at 8:21 PM, Jeff Haran jha...@bytemobile.com wrote:
 Graeme,

 Perhaps, but that's not what I asked about. It seems to me the essence of GPL 
 is that it grants people the right to modify GPL sources like the Linux 
 kernel in any way they want so long as they make those changes available to 
 whoever uses the code in the future. I don't see anything in it that 
 prohibits specific changes. So if I take a symbol that in the sources from 
 kernel.org is declared with EXPORT_SYMBOL_GPL(), make a 1 line change that 
 declares it EXPORT_SYMBOL() and put that on a publically available web site, 
 how have I violated GPL?

 Let's say I then ship a product that uses that custom kernel and a non-GPL 
 kernel module of my own writing that only works with the custom kernel, how 
 is that prohibited in the GPL license?

 Not that I am planning on doing this and I've never done it in the past, but 
 technically it seems that there would be no violation here.

 Thanks,

 Jeff Haran


I assume you know it is against the GPL to remove the license statements.

If I was to write it, the implementation of EXPORT_SYMBOL_GPL() would
have embedded license statements.

Thus if you removed it, you would be removing a license statement and
are in violation of the GPL.

Somehow, I think the kernel legal brains have come up with even better
ideas than I have.

Greg (not KH)

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

RE: GPL-only symbol Error

[Jeff Haran]

 -Original Message-
 From: Guillaume Knispel [mailto:gknis...@proformatique.com]
 Sent: Tuesday, November 22, 2011 5:40 PM
 To: Jeff Haran
 Cc: Graeme Russ; Greg KH; Sengottuvelan S; Kernel Newbies
 Subject: Re: GPL-only symbol Error
 
 On Tue, 22 Nov 2011 17:21:46 -0800
 Jeff Haran jha...@bytemobile.com wrote:
 
   -Original Message-
   From: Graeme Russ [mailto:graeme.r...@gmail.com]
   Sent: Tuesday, November 22, 2011 5:10 PM
   To: Jeff Haran
   Cc: Greg KH; Sengottuvelan S; Kernel Newbies
   Subject: Re: GPL-only symbol Error
  
   Hi Jeff,
  
   On Wed, Nov 23, 2011 at 11:34 AM, Jeff Haran
 jha...@bytemobile.com
   wrote:
-Original Message-
From: Greg KH [mailto:g...@kroah.com]
Sent: Tuesday, November 22, 2011 2:44 PM
To: Jeff Haran; Sengottuvelan S; Kernel Newbies
Subject: Re: GPL-only symbol Error
   
On Tue, Nov 22, 2011 at 02:35:24PM -0800, Jeff Haran wrote:
 I've seen others when faced with this who build their own kernels
from
 sources just modify the problematic EXPORT_SYMBOL_GPL()s to
 EXPORT_SYMBOL()s. I don't know if that is legal. I wouldn't do it
 personally. Consult a lawyer before you go down that road.
   
It is not legal and companies have gotten into big trouble by trying
to
do that, or by creating gpl-condom kernel modules that wrap gpl-
 only
symbols and export them again.  Do not do that without the full buy-in
from your legal department as they do not want to hear about it from
an
external query first.
   
   
Greg,
   
Just curious, can you provide links to these cases?
   
I've read the COPYING file at the top of the Linux source tree. I am not
a lawyer, but I don't see anything in it that would prohibit somebody
from taking the GPL kernel sources, changing the
 EXPORT_SYMBOL_GPL()s
   to
EXPORT_SYMBOL()s, publishing that modified kernel source as required
 by
the GPL license but then keep their module source that uses the now
non-GPL symbols private. It seems like it should be prohibited in the
spirit of open source, but I don't see any mention of these symbol
declarations in the license.
  
   The mere fact that the gpl-condom module links to
 EXPORT_SYMBOL_GPL
   functions make it, in and of itself, a GPL module
  
   ergo, there is no such thing as a gpl-condom module
  
   Regards,
  
   Graeme
 
  Graeme,
 
  Perhaps, but that's not what I asked about. It seems to me the essence of
 GPL is that it grants people the right to modify GPL sources like the Linux
 kernel in any way they want so long as they make those changes available to
 whoever uses the code in the future. I don't see anything in it that prohibits
 specific changes. So if I take a symbol that in the sources from kernel.org is
 declared with EXPORT_SYMBOL_GPL(), make a 1 line change that declares it
 EXPORT_SYMBOL() and put that on a publically available web site, how have I
 violated GPL?
 
  Let's say I then ship a product that uses that custom kernel and a non-GPL
 kernel module of my own writing that only works with the custom kernel,
 how is that prohibited in the GPL license?
 
  Not that I am planning on doing this and I've never done it in the past, but
 technically it seems that there would be no violation here.
 
 I guess you can't really take some GPL code from a third party, do some
 random paperwork or magic trick that basically constitute an unilateral
 declaration from you that what was once considered and clearly
 identified as derivatives are not anymore, publish said derivatives
 that you pretend are not and do so in an incompatible licence, and get
 along with it. On the contrary such behavior would very probably
 constitute something like willfulness infringement if such thing is
 applicable (but IANAL and so over...)

If I planned to do this, I wouldn't change the license. The custom kernel 
sources I posted to the would-be web site would still have the same COPYING 
file.

You say it can't be done, but so far I've read nothing that convinces me that 
it would be illegal according to the verbiage of the license.

Again, given the scenario I described, how would the existing GPL2 license file 
that accompanies the kernel sources prohibit it?

It seems to me that the answer so far is nothing and as such the previous 
assertion that doing this would be illegal is not supported by the law, at 
least not in the US where I live. Perhaps in other countries the laws of 
copyright are different, but if there has been anything in these recent 
international trade agreements that has been successful it has been a 
standardization of the intellectual property law that governs this kind of 
thing across different national boundaries.

I personally am all for open source. I wish I could have been more successful 
in the past at convincing my employers to open up their proprietary modules and 
push them upstream, if no other reason that it would have made my life 

RE: GPL-only symbol Error

[Jeff Haran]

 -Original Message-
 From: Greg Freemyer [mailto:greg.freem...@gmail.com]
 Sent: Tuesday, November 22, 2011 5:57 PM
 To: Jeff Haran
 Cc: Graeme Russ; Greg KH; Sengottuvelan S; Kernel Newbies
 Subject: Re: GPL-only symbol Error
 
 On Tue, Nov 22, 2011 at 8:21 PM, Jeff Haran jha...@bytemobile.com
 wrote:
  Graeme,
 
  Perhaps, but that's not what I asked about. It seems to me the
essence of
 GPL is that it grants people the right to modify GPL sources like the
Linux
 kernel in any way they want so long as they make those changes
available to
 whoever uses the code in the future. I don't see anything in it that
prohibits
 specific changes. So if I take a symbol that in the sources from
kernel.org is
 declared with EXPORT_SYMBOL_GPL(), make a 1 line change that declares
it
 EXPORT_SYMBOL() and put that on a publically available web site, how
have I
 violated GPL?
 
  Let's say I then ship a product that uses that custom kernel and a
non-GPL
 kernel module of my own writing that only works with the custom
kernel,
 how is that prohibited in the GPL license?
 
  Not that I am planning on doing this and I've never done it in the
past, but
 technically it seems that there would be no violation here.
 
  Thanks,
 
  Jeff Haran
 
 
 I assume you know it is against the GPL to remove the license
statements.
 
 If I was to write it, the implementation of EXPORT_SYMBOL_GPL() would
 have embedded license statements.
 
 Thus if you removed it, you would be removing a license statement and
 are in violation of the GPL.

Perhaps, but you didn't write it.

The implementation of EXPORT_SYMBOL_GPL() is this:

/* For every exported symbol, place a struct in the __ksymtab section */
#define __EXPORT_SYMBOL(sym, sec)   \
extern typeof(sym) sym; \
__CRC_SYMBOL(sym, sec)  \
static const char __kstrtab_##sym[] \
__attribute__((section(__ksymtab_strings), aligned(1))) \
= MODULE_SYMBOL_PREFIX #sym;\
static const struct kernel_symbol __ksymtab_##sym   \
__used  \
__attribute__((section(__ksymtab sec), unused))   \
= { (unsigned long)sym, __kstrtab_##sym }

#define EXPORT_SYMBOL_GPL(sym)  \
__EXPORT_SYMBOL(sym, _gpl)

How does that constitute a legal license document? It's C preprocessor
gibberish that only a software engineer can understand. Law is decided
in courts, in juries of peers, at least the US. Any judge that is asked
to rule on the validity of the above as a legal document would laugh the
attorney out of court.
 
 Somehow, I think the kernel legal brains have come up with even better
 ideas than I have.

I would hope they had, but so far I've seen no evidence to that effect.

Jeff Haran




___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: GPL-only symbol Error

[Greg KH]

On Tue, Nov 22, 2011 at 05:21:46PM -0800, Jeff Haran wrote:
 Perhaps, but that's not what I asked about. It seems to me the essence
 of GPL is that it grants people the right to modify GPL sources like
 the Linux kernel in any way they want so long as they make those
 changes available to whoever uses the code in the future. I don't see
 anything in it that prohibits specific changes. So if I take a symbol
 that in the sources from kernel.org is declared with
 EXPORT_SYMBOL_GPL(), make a 1 line change that declares it
 EXPORT_SYMBOL() and put that on a publically available web site, how
 have I violated GPL?

If you distribute that, nothing.  But if you somehow think that
protects your closed source kernel module to now use that symbol
exported by the condom module, you are mistaken.

Again, talk to a lawyer for the details, they are the best ones to
answer this for you, we aren't.

Would you ask a lawyer mailing list for medical questions?

greg k-h

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: GPL-only symbol Error

[Greg KH]

On Tue, Nov 22, 2011 at 06:10:27PM -0800, Jeff Haran wrote:
 But it doesn't do anybody any good to spread misinformation about this
 topic, particularly with regard to what is and isn't legal.

I agree, please don't continue it, but rather, consult a lawyer if you
have further questions.

greg k-h

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: busybox utility

[naveen yadav]

Dear Friends thanks for reply,

I post on Kernel newbies since i could not found any good place to post.
We do not want to use busybox due to its licence and security issue.
So is there any alternate to busybox ..

Thanks


On Wed, Nov 23, 2011 at 12:22 AM, Thomas Petazzoni
thomas.petazz...@free-electrons.com wrote:
 Le Tue, 22 Nov 2011 10:11:26 -0800,
 Jeff Haran jha...@bytemobile.com a écrit :

  This question is not kernel-related, so it shouldn't be posted on
  linux-kernel@ or kernelnewbies@.

 Oh, give me a break. kernelnewbies, as in newbie, as in don't
 know their way around yet.

 It's a valid point that questions like this shouldn't go to
 linux-kernel, but if newbies can't post questions to a newbies
 mailing list without fear of being chastised for it, then I don't
 what the list is for.

 Sorry if my sentence felt offensive. I think the purpose of the newbies
 mailing list is also for newbies to *learn* their way around, and
 explaining that posting a question on the right mailing list is I think
 important. I have no problem with newbies doing mistakes, otherwise I
 wouldn't be subscribed on this list and I wouldn't reply to questions.
 So, really, my statement was really for information, but I understand
 that my wording might have been understood differently. Apologies if
 that was the case.

 Best regards,

 Thomas
 --
 Thomas Petazzoni, Free Electrons
 Kernel, drivers, real-time and embedded Linux
 development, consulting, training and support.
 http://free-electrons.com

 ___
 Kernelnewbies mailing list
 Kernelnewbies@kernelnewbies.org
 http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies


___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: busybox utility

[Mandeep Sandhu]

 I post on Kernel newbies since i could not found any good place to post.
 We do not want to use busybox due to its licence and security issue.

What security issue are you concerned about? Can you explain your
concern a little more.

Also, if you have some busybox specific question, ppl on the busybox
mailing list will be able to help you better.

-mandeep

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

plz help with the serial port pgm....

[shibin k reeny]

rs232 read pgm is not working when i connect in loopback but write is
wroking plz help...

pgm

#include stdio.h
#include stdlib.h
#include string.h
#include unistd.h
#include fcntl.h
#include termios.h

#define _POSIX_SOURCE 1

#define BAUDRATEB9600
#define MODEMDEVICE/dev/ttyS1

int main(void)
{
int ret, fd;
struct termios tio;
char buf[255];

/* Open device for reading and writing */
fd = open(MODEMDEVICE, O_RDWR | O_NOCTTY | O_NDELAY);
if (fd  0) {
fprintf(stderr, Error opening device %s\n, MODEMDEVICE);
goto done;
}
/* get current port attributes */
ret = tcgetattr(fd, tio);
if (ret  0) {
fprintf(stderr, Error retreiving attributes, ret=%d\n, ret);
goto done;
}
tcflush(fd, TCIFLUSH);

/* set port input speed */
ret = cfsetispeed(tio, BAUDRATE);
if (ret) {
fprintf(stderr, Error setting input Baud rate, ret=%d\n, ret);
goto done;
}
/* set port output speed */
ret = cfsetospeed(tio, BAUDRATE);
if (ret) {
fprintf(stderr, Error setting output Baud rate, ret=%d\n, ret);
goto done;
}

tio.c_cflag = ~CRTSCTS; /* HW flow ctl OFF */
tio.c_cflag = ~PARENB; /* no parity */
tio.c_cflag = ~CSTOPB; /* 1 stop bit */
tio.c_cflag = ~CSIZE; /* 1 stop bit */
tio.c_cflag |= CS8; /* char size; 8N1 */
tio.c_cflag |= CREAD; /* enable receiver */
/* set port attributes */
ret = tcsetattr(fd, TCSANOW, tio);
if (ret  0) {
fprintf(stderr, Error setting attributes, ret=%d\n, ret);
goto done;
}
fcntl(fd, F_SETFL, FNDELAY);
ret = write(fd, ab\r, 3);
if (ret  0)
fprintf(stderr, write() of 3 bytes failed!, ret=%d\n, ret);

printf(no of data written = %d\n,ret);
usleep(5000);
while (1) {
ret = read(fd, buf, sizeof(char));
if (ret  0) {
printf(%c,buf[0]);

}
}
done:
close(fd);
return 0;
}


##
output../a.out
no of data written = 3





strace ./a.out
read(3, 0xbfa8d5dd, 1)  = -1 EAGAIN (Resource temporarily
unavailable)
read(3, 0xbfa8d5dd, 1)  = -1 EAGAIN (Resource temporarily
unavailable)
read(3, 0xbfa8d5dd, 1)  = -1 EAGAIN (Resource temporarily
unavailable)
read(3, 0xbfa8d5dd, 1)  = -1 EAGAIN (Resource temporarily
unavailable)
read(3, 0xbfa8d5dd, 1)  = -1 EAGAIN (Resource temporarily
unavailable)
read(3, 0xbfa8d5dd, 1)  = -1 EAGAIN (Resource temporarily
unavailable)
read(3, 0xbfa8d5dd, 1)  = -1 EAGAIN (Resource temporarily
unavailable)
read(3, 0xbfa8d5dd, 1)  = -1 EAGAIN (Resource temporarily
unavailable)
read(3, 0xbfa8d5dd, 1)  = -1 EAGAIN (Resource temporarily
unavailable)
___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies

Re: plz help with the serial port pgm....

[Dave Hylands]

Hi shibin

On Tue, Nov 22, 2011 at 9:48 PM, shibin k reeny shibinkre...@gmail.com wrote:
 rs232 read pgm is not working when i connect in loopback but write is
 wroking plz help...

Rather than trying to debug your program, I'll point you at this document:
http://tldp.org/HOWTO/Serial-Programming-HOWTO/

and point at some sample code which I wrote a while ago, which I know works:
http://svn.hylands.org/host/sertest/sertest.c

-- 
Dave Hylands
Shuswap, BC, Canada
http://www.davehylands.com

___
Kernelnewbies mailing list
Kernelnewbies@kernelnewbies.org
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies