May 2006 - openSUSE Programming - openSUSE Mailing Lists

Possible bug: boost.signals with '-pthread'?
by Pavel Syomin 23 May '06

23 May '06

Hi, all! I am novice in this list, so don't beat freely :) I confronted with a strange thing using boost.signals on SLES 9 SP3. I suspect, that it is bug, becase on FC4 and RHEL AS 3 all fine. Test case: --------------------- CUT HERE ------------------ #include <boost/signal.hpp> #include <iostream> struct MySlot { void operator() () { std::cout << "YES!" << std::endl; } }; MySlot myslot; boost::signal<void ()> mysig; int main() { mysig.connect(myslot); return 0; } --------------------- CUT HERE ------------------ Compile and run: c++ test.cc -o test -lboost_signals ./test Programm will succesfully exit. Compile and run: c++ test.cc -o test -lboost_signals -pthread ./test Programm will not exit never. Any ideas? P.S. I can't use GCC without '-pthread' because my program used thread...

1 0

How to use epoll and poll
by Chaitanya Chalasani 16 May '06

16 May '06

Hi, I am using SuSE linux 10.0. To impove performace of network I/O we would like to use poll or epoll. Is it neccesary that /dev/poll and /dev/epoll is required. If so how do we create that devices? Thanks in advance -- Chaitanya Chalasani ------------------- "The ultimate truths of nature are so subtle that they can be revealed only under immense silence"

2 3

debugging a 32-bit code by DDD on a 64-bit laptop
by Maura Edeweiss Monville 16 May '06

16 May '06

I compiled one routine (m15gui.c) belonging to a big Monte Carlo package with compiler option -g to build the symbol table just for this routine only. It sounds confusing. But what I did is to append the routine source file "m15gui.c" to the GNUmakefile as follows: SRCS := marsmain.f m1505.f m5gui.c Then I replaced the original compiler and linker options: CFLAGS := -O3 -xarch=v8 LDFLAGS := -fast -xarch=v8 -xlang=f77 -L/usr/openwin/lib with the following: CFLAGS := -g -xarch=v8 LDFLAGS := -fast -g -C -xs -xildoff -xarch=v8 -xlang=f77 -L/usr/openwin/lib It compiled fine. The program I wish to debug resides in /home/mauede/mars15/dev When I launch DDD and open the program a message window pops up. The title is: DDD No Source The message reads as follows: /home/mauede/mars15/dev/sysdeps/i386/elf/start.S: no such file or directory OK Help I swear there is no such directory: sysdeps/i386/elf in my working directory: /home/mauede/mars15/dev Please, notice that my laptop is 64-bit and the message above sounds like it's searching for some 32-bit feature ... ? I'm debugging a 32-bit code....... Are the DDD settings to be changed ? If this is the case then how shall I change them ??? Isn't "elf" something system related ? Thank you so much for your help, -- Maura E.M B

2 2

HowTo create multiple profiles
by Maura Edeweiss Monville 15 May '06

15 May '06

When I joined FNAL for the first couple of weeks I just used my laptop. We could configure both the wire and wireless network cards to work with FNAL network thanks to the help of SuSE mailing-list experts. Then I got an office and a desk computer at work. At home I got the DSL activated and thanks to a AT&T technician I could configure both my laptop wire and wireless cards to work with the AT&T network. This morning I came to the lab with my laptop as my office has been remodeled and the desktop computer is still disconnected. As you can imagine I could not connect my laptop to FNAL network either through the wire or the wireless card. It is the same problem that arouse with the printers. I configured my laptop printing system to communicate with CUPS server here at FNAL. So I can print from my laptop when I'm at the lab. But I can no more print at home as my local EPSON ink-jet printer configuration has been overwritten or invalidated by the configuration I had to do for printing at the lab. Some time ago someone told me it's possible to have a profile for each environment where the laptop is operated. My understanding is that each profile uniquely identifies some environment specific settings so that different environment configurations can be used alternatively. How can I create and use more profiles ? How can I relate the single profile with a specific configuration ? Thank you so much for your help, -- Maura E.MA

2 1

HowTo set DDD to debug a 32-bit code on a 64-bit laptop
by Maura Edeweiss Monville 15 May '06

15 May '06

I compiled one routine (m15gui.c) belonging to a big Monte Carlo package with compiler option -g to build the symbol table just for this routine only. It sounds confusing. But what I did is to append the routine source file "m15gui.c" to the GNUmakefile as follows: SRCS := marsmain.f m1505.f m5gui.c Then I replaced the original compiler and linker options: CFLAGS := -O3 -xarch=v8 LDFLAGS := -fast -xarch=v8 -xlang=f77 -L/usr/openwin/lib with the following: CFLAGS := -g -xarch=v8 LDFLAGS := -fast -g -C -xs -xildoff -xarch=v8 -xlang=f77 -L/usr/openwin/lib It compiled fine. The program I wish to debug resides in /home/mauede/mars15/dev When I launch DDD and open the program a message window pops up. The title is: DDD No Source The message reads as follows: /home/mauede/mars15/dev/sysdeps/i386/elf/start.S: no such file or directory OK Help Please, notice that my laptop is 64-bit and the message above sounds likeit's searching for some 32-bit feature ... ? I'm debugging a 32-bit code....... Are the DDD settings to be changed ? If this is the case then how shall I change them ??? Thank you so much for your help, -- Maura E.M

1 0

installing mono on opensuse 10.0
by Ferenc Stelcz 12 May '06

12 May '06

Dear List, I'm a newbie in .net and C# programming and read a lot about the mono project and I felt now it is time for me to learn a bit of it. :) So I have an up-to-date opensuse 10.0 and installed the MONO/CLR packages from yast. But if I want to start monodevelop i get the following error message: /opt/gnome/lib/libgnomeui/gnome_segv: symbol lookup error: /opt/gnome/lib64/libgtk-x11-2.0.so.0: undefined symbol: g_object_compat_control I searched the web but found no mono specific answer instead there are only gnome related answer saying I should up/downgrade glib. The current version of glib2 by me is 2.8.5-18 and glib is 1.2.10-606. Did somebody get an error like this? If yes then how to solve it, if no then where shall I start finding the solution? Best regards Ferenc Stelcz p.s. Tried with the binary installer from go-mono.com but that fails too complaining about gnome-sharp imports. Then tried with Red Carpet but because of a server misconfiguration the download directory on the server gives HTTP 403.

1 0

threads and file descriptors
by Per Jessen 10 May '06

10 May '06

I'm seeing a fairly reproducable problem in which the same file descriptor is being used concurrently in two threads, but _unintentionally_. I understand that open file descriptors are shared by all threads, but somehow it looks like calling socket() in one thread is able to return a filedescriptor that's already in use by another thread. Here's an example (a little complicated, but bear with me): this is just stderr debugging output from my app which is running 3 threads at this time. The output is in chronological order. The thread# is in [] to the left, my comments are prefixed with >>> : [01] using fd 16 for /tmp/ibwd.1 [01] state-transition: 2->3 [01] writing to left ( 0 bytes left): 250 Ok: queued as 29C9E4637 [01] state 3; left event LEFT_RSET: [6 bytes] RSET [01] all done <message-id>: 2995 bytes [01] writing to right ( 0 bytes left): RSET [01] state-transition: 3->1 [01] state 1; right event RIGHT_354: [37 bytes] 354 [01] using fd 14 for /tmp/ibwd.1 >>> above thread#1 has opened a file and gets fd 14 for it. [01] state-transition: 1->2 [02] crm114[25958, 9 categories] stage done, 0.509 seconds. [02] connected to juggernautd(14) at localhost:787 >>> above thread#2 opens a tcp connection to another app, and ALSO uses fd 14. [01] state 2; left event LEFT_EOM: [3 bytes] . write_string(): unable to write to 14; error 9: Bad file descriptor >>> above thread#1 chokes when trying to write to fd 14. Each thread is independent of the others, except during allocation of work. Are there any special precautions I need to take when doing IO in a multi-threaded environment? /Per Jessen, Zürich

4 28

RE: [suse-programming-e] SPAM: Question About Driver Update Floppy
by Black, Alain 08 May '06

08 May '06

The suse-sles-e(a)suse.com mailing list will probably be more responsive to this question. > -----Original Message----- > From: McArthor [mailto:macarthor@eyou.com] > Sent: Sunday, May 07, 2006 10:16 PM > To: suse-programming-e(a)suse.com > Subject: [suse-programming-e] SPAM: Question About Driver Update Floppy > > Hi all, > i've made a NIC (Broadcom NIC) driver upgrade floppy disk containing > default/smp/bigsmp .ko files. > I added update.post script to this floppy disk. This update.post aims to > replace OS-embedded tg3 NIC driver (tg3.ko) with my own Broadcom NIC > driver (bcm5700.ko), and without the end customer to do anything after > installation. > My question is: What should be included in update.post to modify the SUSE > to use my broadcom.ko for default NIC driver, instead of OS-embedded > tg3.ko. > Many thanks for your looking-inside. > > > > > > > > --最新热品抢鲜体验! > -- > http://freemail5.eyou.com/sys_sig.php?url=aHR0cDovL2V5b3VhZnAuYWxseWVzLmNv > bS9tYWluL2FkZmNsaWNrP2RiPWV5b3VhZnAmYmlkPTk1LDgxLDMmY2lkPTczLDksMSZzaWQ9MT > Y4JmFkdmlkPTQmY2FtaWQ9NCZzaG93PWlnbm9yZSZ1cmw9aHR0cDovL2FkZmFybS5tZWRpYXBs > ZXguY29tL2FkL2NrLzQwODAtMjI5MDMtOTQ5OS0wP2FpZD1leW91O3RleHQ7MTAzJm1wcm89aH > R0cDovL3BhZ2VzLmViYXkuY29tLmNuL25ld2l0ZW0vaW5kZXguaHRtbA== > > --http://sms.eyou.com > --无忧二二族、足球大富翁...尽在亿邮短信

1 0

SPAM: Question About Driver Update Floppy
by McArthor 08 May '06

08 May '06

Hi all, i've made a NIC (Broadcom NIC) driver upgrade floppy disk containing default/smp/bigsmp .ko files. I added update.post script to this floppy disk. This update.post aims to replace OS-embedded tg3 NIC driver (tg3.ko) with my own Broadcom NIC driver (bcm5700.ko), and without the end customer to do anything after installation. My question is: What should be included in update.post to modify the SUSE to use my broadcom.ko for default NIC driver, instead of OS-embedded tg3.ko. Many thanks for your looking-inside. --最新热品抢鲜体验! --http://freemail5.eyou.com/sys_sig.php?url=aHR0cDovL2V5b3VhZnAuYWxseWVzLmNvbS9tYWluL2FkZmNsaWNrP2RiPWV5b3VhZnAmYmlkPTk1LDgxLDMmY2lkPTczLDksMSZzaWQ9MTY4JmFkdmlkPTQmY2FtaWQ9NCZzaG93PWlnbm9yZSZ1cmw9aHR0cDovL2FkZmFybS5tZWRpYXBsZXguY29tL2FkL2NrLzQwODAtMjI5MDMtOTQ5OS0wP2FpZD1leW91O3RleHQ7MTAzJm1wcm89aHR0cDovL3BhZ2VzLmViYXkuY29tLmNuL25ld2l0ZW0vaW5kZXguaHRtbA== --http://sms.eyou.com --无忧二二族、足球大富翁...尽在亿邮短信

1 0

should I use memmove instead of strcpy just to appease valgrind?
by Per Jessen 07 May '06

07 May '06

valgrind always complains about overlapping strcpy()s, even when the src-address is clearly > dst-address. I know the strcpy manpage says the operands may not overlap, but surely an overlap where src>dst is still ok? Why is valgrind unhappy about it? /Per Jessen, Zürich

3 7