[uml-devel] Testing needed for terminal handling patches

[uml-devel] Testing needed for terminal handling patches

[Jeff Dike]

I've posted a set of patches that fiddle with UML's handling of terminal modes
to http://user-mode-linux.sourceforge.net/patches.html

I'd like people to try them out and report bad experiences, especially if you
saw the 'netdevice 0' hang or stairstepping output during boot.  They are the
catch-eintr, bb-raw, and clean-raw patches.

In particular, I removed the complain argument to raw, which means that it
will always complain if something failed.  The reason it was there in the
first place is that a long, long time ago, in a universe far, far away, I
saw a failure there on stdout call printk, which would recurse through the
same code, causing the same failure, reporting it through printk again, ad
infinitum.  I'm curious as to whether this infinite recursion was sparked
by an EINTR or something else.

Also, TCSACRAIN is always used instead of TCSANOW.  I don't see any reason
that I need NOW instead of DRAIN.

				Jeff



-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] Testing needed for terminal handling patches

[Allen Chan]

I got a number of errors compiling the kernel with the new patches.   
The log of the compile at the point of the first three errors are as 
follows:

-----------------
gcc   -Wall -Wstrict-prototypes -Wno-trigraphs -O2 
-fno-strict-aliasing -fno-common -U__i386__ -Ui386 -DUM_FASTCALL -g 
-D__arch_um__ -DSUBARCH=\"i386\" -D_LARGEFILE64_SOURCE  
-I/home/alydar/play/kernel/linux-2.4.26/arch/um/include 
-I/home/alydar/play/kernel/linux-2.4.26/arch/um/kernel/tt/include 
-I/home/alydar/play/kernel/linux-2.4.26/arch/um/kernel/skas/include 
-D_GNU_SOURCE -c -o process.o process.c
process.c: In function `handle_trap':
process.c:64: warning: use of conditional expressions as lvalues is 
deprecated
process.c:80: error: parse error before "while"
...
-----------------
gcc   -Wall -Wstrict-prototypes -Wno-trigraphs -O2 
-fno-strict-aliasing -fno-common -U__i386__ -Ui386 -DUM_FASTCALL -g 
-D__arch_um__ -DSUBARCH=\"i386\" -D_LARGEFILE64_SOURCE  
-I/home/alydar/play/kernel/linux-2.4.26/arch/um/include 
-I/home/alydar/play/kernel/linux-2.4.26/arch/um/kernel/tt/include 
-I/home/alydar/play/kernel/linux-2.4.26/arch/um/kernel/skas/include 
-D_GNU_SOURCE -c -o sigio_user.o sigio_user.c
sigio_user.c: In function `check_one_sigio':
sigio_user.c:72: error: too few arguments to function `__raw'
...
-----------------
gcc   -Wall -Wstrict-prototypes -Wno-trigraphs -O2 
-fno-strict-aliasing -fno-common -U__i386__ -Ui386 -DUM_FASTCALL -g 
-D__arch_um__ -DSUBARCH=\"i386\" -D_LARGEFILE64_SOURCE  
-I/home/alydar/play/kernel/linux-2.4.26/arch/um/include 
-I/home/alydar/play/kernel/linux-2.4.26/arch/um/kernel/tt/include 
-I/home/alydar/play/kernel/linux-2.4.26/arch/um/kernel/skas/include 
-D_GNU_SOURCE -c -o user_util.o user_util.c
user_util.c:122:29: macro "raw" passed 2 arguments, but takes just 1
...
-----------------


I was able to fix the typo on the first one by modifying line 80 of 
arch/um/kernel/skas/process.c to:
    CATCH_EINTR(err = waitpid(pid, &status, WUNTRACED));
from:
    err = CATCH_EINTR(waitpid(pid, &status, WUNTRACED));

However, it is not clear to me how the raw() and __raw() functions and 
macros were intended to be handled on the other errors.  I have a 
strange suspicion that the two "raw" patches posted on the 
incremental patches site are incomplete.


On Tuesday 10 August 2004 01:15 am, Jeff Dike wrote:
> I've posted a set of patches that fiddle with UML's handling of
> terminal modes to
> http://user-mode-linux.sourceforge.net/patches.html
>
> I'd like people to try them out and report bad experiences,
> especially if you saw the 'netdevice 0' hang or stairstepping
> output during boot.  They are the catch-eintr, bb-raw, and
> clean-raw patches.


-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] Testing needed for terminal handling patches

[Jeff Dike]

alydar@users.sourceforge.net said:
> I got a number of errors compiling the kernel with the new patches.
> The log of the compile at the point of the first three errors are as
> follows: 

These should be fixed.  I just ran builds of each patch, starting from a clean
pool.

				Jeff



-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel

Re: [uml-devel] Testing needed for terminal handling patches

[Matt Clay]

Jeff Dike wrote:
> 
> I'd like people to try them out and report bad experiences, especially if you
> saw the 'netdevice 0' hang or stairstepping output during boot.  They are the
> catch-eintr, bb-raw, and clean-raw patches.

Looks good, no more stairstepping during boot. :)

  - Matt


-------------------------------------------------------
SF.Net email is sponsored by Shop4tech.com-Lowest price on Blank Media
100pk Sonic DVD-R 4x for only $29 -100pk Sonic DVD+R for only $33
Save 50% off Retail on Ink & Toner - Free Shipping and Free Gift.
http://www.shop4tech.com/z/Inkjet_Cartridges/9_108_r285
_______________________________________________
User-mode-linux-devel mailing list
User-mode-linux-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/user-mode-linux-devel