Hi, Sergio.
It is really simple. I think I just left out too many details, believing that everyone was as disgustingly familiar with the svgalib compile process as I have unfortunately become. I wish I were far less familiar with it ;)
When svgalib is compiled with the 'make' command, it creates svgalib as well as svgalibhelper, all as a part of the same 'make' command. As I stated, 'make' failed on creating svgalib_helper. The sequence of events in the compile operation is such that svgalib had already been created by this time. By setting the NOHELPER option, svgalib itself, which was created OK, runs with no need for svgalib_helper. Thus the failure to create svgalib_helper during the 'make' operation is of no consequence, as I had indicated.
As far as Linrad goes, compiling linrad-01.35 failed with the error I noted. I therefore of course couldn't run it.
But Linrad-02.03 compiled fine with no error. There is an OSS problem with the kernel 2.6.15.1 that I am trying to work out that causes anything that used OSS to lock up, and Linrad is of course no exception here. But this is not a Linrad issue. I was stupid, and SHOULD have tried the 'stock' audio to see if in ETCH it will work at 96KHz sampling before trying OSS. But I did not, because of my own stupidity. I will next reinstall the original stock audio and see if that works. If it does I will not bother with OSS. I am always frustrated with the problems OSS has each time the kernel is updated, but so far it has been the only solution that has allowed me to sample at 96000 Hz with no problems. If the stock audio ever does the same, OSS will be history ;)
The bottom line is that I always tell others in the strongest terms, "Don't update your kernel unless you have a really good reason to do it." But I cannot help myself, and always want to see what will happen. Do not try this at home ;)
73,
Roger
W3SZ
-------------- Original message ----------------------
From: Sergio <ik2mmb@xxxxxxxx>
Roger,
I am a Debian user so I'm interested yet I am a bit confused. You say
you got several errors on the make which also seems not to conclude
properly; how can you possibly install it ok after? Or for 'debian new
version' you intend the testing and not the etch version. Sorry if I am
maybe misinterpreting something but maybe you can clarify for me.
Thanks
Sergio ik2mmb
w3sz wrote:
I am testing Debian Etch [testing] with
OSS version: 3.99.4a
OSS build: 200603251735
Kernel version: Linux debian 2.6.15-1-686-smp #2 SMP Mon Mar 6 15:34:50
UTC 2006 i686 GNU/Linux
Kernel vermagic: 2.6.15-1-686-smp SMP 686 gcc-4.0
Modutils version: 3.2.2
Also 4.0.3-1 and nasm 0.98.38.
I find when trying to 'make' Linrad-01.35 that I get the following error:
main.o ui.o setvga.o setad.o fft1.o buf.o mouse.o wide_graph.o \
getiq.o fft0.o rxtest.o hires_graph.o fftasm.o baseb_graph.o mix2.o \
afc_graph.o hwaredriver.o sellim.o palette.o gifsave.o timing.o help.o \
calibrate.o calsub.o calsub2.o llsqvar.o llsq.o caliq.o timf2.o wcw.o \
powtim.o selvar.o fft2var.o sigvar.o screenvar.o uivar.o fft1var.o \
fft3var.o calvar.o blnkvar.o mix1.o fft1_re.o fft1_approx.o
fft1_complex.o \
simdasm.o blank1.o fft2.o pol_graph.o split.o afcsub.o modesub.o spur.o \
cohsub.o coherent.o eme.o spursub.o freq_control.o fft2mmxb.o fft2mmxa.o \
tune.o txtest.o network.o fft2mmxc.o hware.o morse.o coh_osc.o cwspeed.o \
netvar.o settx.o cwdetect.o txvar.o vlab.o fft3.o
/usr/bin/ld: unrecognized option '--crefmain.o'
/usr/bin/ld: use the --help option for usage information
collect2: ld returned 1 exit status
make: *** [linrad] Error 1
---------------end of make error
Linrad-02.03 installs fine without this error under the new Debian
distro and kernel!
This is just an FYI.
Since I am sending this via the list I will post a couple of other
interesting things.
I am playing with installing Debian testing [Etch] using Netinst, from
http://www.debian.org/devel/debian-installer/
This installed OK except that I had to go in and set the network
parameters. For some reason netinst assumed DHCP, so I had to enter
the dnsservers and ip address/netmask/gateway stuff after install.
svgalib install ended with an error, with svgalib_helper failing to
install [of course]. Setting the NOHELPER option rendered this problem
moot. I again had to use the mdev /dev/input/mice statement to get the
mouse to work in svgalib. I had to use the VESA driver with my Intel
i810 video to get decent video resolution choices in svgalib.
OSS remains a problem. It will install and I can turn it on with
./soundon and configure it with ./soundconf, but when it is used either
via ./osstest or ./linrad, it will lock up the computer thoroughly so
that I must hold in the ac power button for about 10 seconds to shut
off the computer and then reboot. cntl-alt-del, cntl-alt-F7,
cntl-alt-F8, escape, cntl-C, etc have no effect in this situation.
To get OSS to install I had to use my usual trick of renaming
soundcore.ko to soundcore.ko.old to stop the sound from installing so
that OSS could install.
I would recommend staying with the stable form of Debian at present
[Sarge]. The most recent kernel I have gotten to work with Sarge is
2.6.12.x. Trying 2.6.14.2 got me into quite a bit of difficulty, so I
decided to just go all the way to Etch and 2.6.15.1 and see if I could
get everything working.
73,
Roger
W3SZ