l4-hackers November 1998

l4-hackers@os.inf.tu-dresden.de

17 participants
29 discussions

Re: L4/Linux problem
by Michael Hohmuth 30 Nov '98

30 Nov '98

"Sun-Mi You" <smyou(a)postech.ac.kr> writes: > I've change Makefile and Linux was compiled sucessfully. > But, the problem still occurs at same point. > > Is there other suggestion? Hmm... Are you sure you re-built the emulib? If in doubt, please "make clean" and rebuild the whole kernel. Changing the Makefile and just doing a "make" in the top-level directory is not enough because the Makefiles are probably not included in the object dependencies. Sorry for being so overly accurate with this; I just want to make sure it's not that known problem. Michael -- hohmuth(a)innocent.com, hohmuth(a)sax.de http://www.sax.de/~hohmuth/

1 0

Re: page size assertion
by Michael Hohmuth 30 Nov '98

30 Nov '98

Eiji Ota <age(a)yk.fujitsu.co.jp> writes: > > > > I'm interested in this fantastic work(of course, L4). When I tried to > > > > feel it, I faced an assertion failed by assert(size == PAGE_SIZE) in > > > > space.cc(178 line). > > > > > > > > The wrong page size seemed to be passed by fpage_map in ipc_map.cc as > > > > a SUPERPAGE_SIZE, but I don't find the root cause for this trouble. > > Hi, after I changed the 113-line in ipc_map.cc to the following, it > worked, though I don't know whether it was correct or not. > > if (size > snd_size // want to send less that a superpage? > | > V > if (size >= snd_size // want to send less that a superpage? This is not the correct fix, I'm afraid. What you do here is to effectively prohibiting the passing of superpage (4-MByte) mappings. I think I have a theory on what the problem in Fiasco is (more on that below). However, it would still be useful to know what the system is trying to do when the error occurs; this would help me to reproduce and track down the problem. My theory is that we fail to test rigorously enough that a receiver really has enough space in its virtual address space to take a superpage mapping. I think we currently test that just 4K are unmapped in the receiver (that is, no mapping exists for the target virtual address); that's not enough. I will try to verify this theory when I'm in the office later today. Michael -- hohmuth(a)innocent.com, hohmuth(a)sax.de http://www.sax.de/~hohmuth/

1 0

Re: page size assertion
by Eiji Ota 30 Nov '98

30 Nov '98

> > > I'm interested in this fantastic work(of course, L4). When I tried to > > > feel it, I faced an assertion failed by assert(size == PAGE_SIZE) in > > > space.cc(178 line). > > > > > > The wrong page size seemed to be passed by fpage_map in ipc_map.cc as > > > a SUPERPAGE_SIZE, but I don't find the root cause for this trouble. Hi, after I changed the 113-line in ipc_map.cc to the following, it worked, though I don't know whether it was correct or not. if (size > snd_size // want to send less that a superpage? | V if (size >= snd_size // want to send less that a superpage? - Eiji

1 0

Re: L4/Linux problem
by Sun-Mi You 30 Nov '98

30 Nov '98

: :I think this is a problem one other user also experienced: The system :call IPC library (called ``emulation library'' for hysterical reasons) :in linux/arch/l4-i386/emulib/ has been linked to a wrong virtual :address because the magic in this directory's Makefile failed. One :possible reason is that you don't have a `cpp' in your $PATH. : :Does it help to replace "cpp" with "gcc -E" in this Makefile? : :Michael :-- I've change Makefile and Linux was compiled sucessfully. But, the problem still occurs at same point. Is there other suggestion?

1 0

ADV: LIVE Adult Paradise -- FREE !!
by Steve 29 Nov '98

29 Nov '98

------------------------------------------------------------- This message is sent in compliance of the new email bill section 301. Per Section 301, Paragraph (a)(2)(C) of S. 1618, further transmissions to you by the sender of this email may be stopped at no cost to you by emailing mailto:sgarner@tophosting.com Please type "Remove" in the subject line. Your request to be remove will be processed within 24 hours. Please DO NOT reply to this email if you wish to be removed. ------------------------------------------------------------- FREE Adult Offer -- Five FREE Minutes -- All LIVE NEED A LITTLE SPICE IN YOUR LIFE? All Your Desires Are Satisfied -- The Best LIVE CyberNet Models Do As You Please! ALL LIVE -- ALL DAY -- ALL NIGHT Seductive Women: http://204.248.170.151/196/index.html HardBody Hunks: http://204.248.170.151/215/index.html ** LIVE 1-on-1 Shows! ** MALE or FEMALE! ** 5 FREE MINUTES of HOT ADULT LIVE ACTION! ** SATISFACTION GUARANTEED! All our shows are 100% LIVE and INTERACTIVE for your enjoyment! Ask And You Shall Receive SEDUTIVE WOMEN That Do It ALL! http://204.248.170.151/196/index.html -- OR -- HARDBODY HUNKS Give You What You Ask! http://204.248.170.151/215/index.html <a href="http://204.248.170.151/196/index.html">AOL Users Click Here!</a> See for yourself - We'll give you 5 FREE Steamy Minutes -- Satisfaction Guaranteed!

1 0

Re: page size assertion
by Eiji Ota 28 Nov '98

28 Nov '98

> > I'm interested in this fantastic work(of course, L4). When I tried to > > feel it, I faced an assertion failed by assert(size == PAGE_SIZE) in > > space.cc(178 line). > > > > The wrong page size seemed to be passed by fpage_map in ipc_map.cc as > > a SUPERPAGE_SIZE, but I don't find the root cause for this trouble. > > > > Any ideas ? > > If you can attach a remote GDB via a serial line, you can generate a > backtrace. I'm not familiar with this analysis, anyway, I'll try this, however, I'm happy if you tell me how-to at this point if possible. > This might be a bug either in Fiasco or in an application (in which > case the code in v_insert() must be made more robust; that is, it must > return an error instead of panicking the kernel). > > What are you doing when the assertion fails? Booting? Starting X11? It was booting. I'll dig this out later as far as I can. - Eiji

1 0

Re: page size assertion
by Michael Hohmuth 28 Nov '98

28 Nov '98

Eiji Ota <age(a)yk.fujitsu.co.jp> writes: > I'm interested in this fantastic work(of course, L4). When I tried to > feel it, I faced an assertion failed by assert(size == PAGE_SIZE) in > space.cc(178 line). > > The wrong page size seemed to be passed by fpage_map in ipc_map.cc as > a SUPERPAGE_SIZE, but I don't find the root cause for this trouble. > > Any ideas ? If you can attach a remote GDB via a serial line, you can generate a backtrace. This might be a bug either in Fiasco or in an application (in which case the code in v_insert() must be made more robust; that is, it must return an error instead of panicking the kernel). What are you doing when the assertion fails? Booting? Starting X11? Michael -- hohmuth(a)innocent.com, hohmuth(a)inf.tu-dresden.de http://home.pages.de/~hohmuth/

1 0

page size assertion
by Eiji Ota 27 Nov '98

27 Nov '98

Hi all, I'm interested in this fantastic work(of course, L4). When I tried to feel it, I faced an assertion failed by assert(size == PAGE_SIZE) in space.cc(178 line). The wrong page size seemed to be passed by fpage_map in ipc_map.cc as a SUPERPAGE_SIZE, but I don't find the root cause for this trouble. Any ideas ? - Eiji Ota

1 0

Re: L4-Linux
by Michael Hohmuth 27 Nov '98

27 Nov '98

Harald Fritzsche <harald.fritzsche(a)mpd.de> writes: > i have applied Michaels patch and fortunatly the message > <KERNEL: no page fault handler ...> > does never occur. > > L4Linux boots now. > > Thanks so far and congratulation for such a system (Linux on Fiasco). Congratulations for getting it running, and thanks for going through all the hassles! > startx or something else (xdm,kdm,X,...) scrambles the screen and blocks the > keyboard. I assume that this is causes by the disabled serial interface and > related IRQ's (i have configured out serial interface)? It might as well be a bug in Fiasco. So far, we also haven't been very successful in getting up X11 (I think I've seen it running once or twice, but it must have been a lucky day). Michael -- hohmuth(a)innocent.com, hohmuth(a)sax.de http://www.sax.de/~hohmuth/

1 0

Re: L4/Linux problem
by Michael Hohmuth 27 Nov '98

27 Nov '98

"Sun-Mi You" <smyou(a)postech.ac.kr> writes: > Our L4/Linux was hanged at follow code. > > in 'linux/arch/l4-i386/chead.c ' > > /* create idt */ > { > extern long idt_table; > struct entry_t {unsigned short w0, w1,w2,w3;} ; > struct entry_t *entries = > (struct entry_t *)((unsigned long)emu_lib_text + > ((unsigned long)&idt_table - EMULIB_CODE_ADDRESS)); > unsigned short int tmp; > int i; > > => this point! > > for (i=0; i<20; i++, entries++) > { > tmp = entries->w1; > entries->w1 = entries->w3; > entries->w3 = tmp; > } > > What problem is this? > > ( I think it may be some memory problem. ) I think this is a problem one other user also experienced: The system call IPC library (called ``emulation library'' for hysterical reasons) in linux/arch/l4-i386/emulib/ has been linked to a wrong virtual address because the magic in this directory's Makefile failed. One possible reason is that you don't have a `cpp' in your $PATH. Does it help to replace "cpp" with "gcc -E" in this Makefile? Michael -- hohmuth(a)innocent.com, hohmuth(a)sax.de http://www.sax.de/~hohmuth/

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

l4-hackers November 1998