Bug#1891: bug report
Packages: base The /etc/magic file does not allow the file command to identify elf .a files from aout .a files very easily. Indeed, an elf archive : archive but an aout archive : random archive Would someone update this to something more meaningful? Thanks, Gordon
Bug#4406: Problems with abuse
Package: abuse Version: 1.10-4 Great game, but I am having quite a few problems with its configuration. (1) X-Version Runs ok without options. Instructions suggest to use -grap_pointer -2 -size 640 400 The program hates this, and says something like "grabbing pointer... pausing for -2 seconds" and then hangs up. (2) X-Version Cannot save the game. Falls over with an error message about writing gamma.lsp. Tries setting the umask. I don't know where it is trying to write to, but my guess must be a permissions problem (3) Console Version abuse should have a dependency on aout-svgalib, as this is needed to get it to execute from the console. Maybe to seperate .deb files would be better, one for X and the other for console execution? That way the dependencies could be kept relatively clean. (4) Console Version. Has anyone ever got this version to run? It dies on me complaining about the mouse. It asks me to set MOUSE_TYPE is I have no MS compatible mouse, so; (i) If I use a MS mouse and set the conf file in etc/vga correctly it says that it could not find the mouse driver. (ii)If I set MOUSE_TYPE then it either says that the mouse type is unrecognised (I tries a MouseSystems), or that it is recognised but that the mouse driver is not present. What is wrong here is anyones guess... I am using the standard buzz-fixed distribution. Thanks Gordon PS Great game! (I can still play it under X if I have a small screen and don't save, but it is hard!)
Bug#4407: aout-svgalib library not detected by ldconfig
Package: aout-svgalib Version: 1.28-6 This package installs its libraries in /usr/i486-linuxaout/lib But all the other aout libraries seem to be in /usr/lib/i486-linuxaout/lib Unfortunately /usr/i486-linuxaout/lib is not present by default in /etc/ld.so.conf, and so the libraries cannot been found. As a fix I added it to ld.so.conf and reran "ldconfig -v". I am using the standard buzz-fixed distribution. Thanks Gordon
Bug#4408: buzz-fixed installation disk broken
Package: boot1440-1 Version: buzz-fixed I need to use the custom boot disks (custom no 1) to start my installation process. However, the modules.tgz file is broken (unexpected EOF during the detar process). I looked at the other boot disks, but all the modules.tgz files seem to be in the same position. What it means is that you do not get any of the ethernet modules with the installation process. The fix for me was complicated; I obtained an image*.deb file, used pkzip on another machine to split it up so that I could get it onto floppies, ran win95 and used pkunzip on the destination machine to extract the .deb file, switched to debian, and used dpkg to extract the modules.tgz file. Then I could get my ethernet card running and continue the installation by ftp. This was using the buzz-fixed installation disks, and as I said custom boot no 1. Thanks Gordon
Bug#4409: Problems with the post-install
Package: kernel-image Version: 2.0.5-1 The post-install seems to look for /vmlinuz, and since I don't have one of those it dies off rather quickly! All my kernels are in the /boot directory, as this is a partition on my IDE drive, and / is on a SCSI drive (which would be difficult to boot from). Surely this is not an unusual setup? This used to be the standard method a few years ago (I have been using linux for many years, and debian for a year or so). As a fix, why not first look in /, and then in /boot? I am using the buzz-fixed distribution. Thanks Gordon
Bug#4413: setserial should not use Pre-Depends
> Package: setserial > Version: 2.10-8 > > setserial should not be an essential base package, so it should not > use Pre-Depends. > > Ian. > I have just checked out the bugs web site, and notice that I have quite a few outstanding reports on setserial. I will try and update it, but a lot of changes seem to have been made to the administration of the packages. Is there a new guide to the format of the .deb files and their submission? Thanks, Gordon
Bug#3762: etc/rc.boot/0setserial usually hangs machine
I have read your bug report regarding setserial... are you still having problems? As far as I know, no other person is having similar problems setserial... If you are still having problems, could you get back to me within the next few days, otherwise I will consider the bug report closed. Thanks in advance, Gordon
Bug#4406: Problems with abuse
>> -grab_pointer -2 -size 640 400 >Try just -2 -size 640 400 This works, but does not grab the pointer... >> (2) X-Version >> Cannot save the game. Falls over with an error message >> about writing gamma.lsp. Tries setting the umask. I don't >> know where it is trying to write to, but my guess must be a >> permissions problem >I'll take a look at this. Running the program as root solves this problem. On inspection, the abuse.x version is SGID root, which should instead probably be SUID root. This seems to solve my problems. Too bad the games and gamma.lsp files are not saved to the user's home directory instead, but then I don't expect you have the source code to make this possible ;-). >> (3) Console Version >> abuse should have a dependency on aout-svgalib, as this is needed to >> get it to execute from the console. Maybe to seperate .deb files would >> be better, one for X and the other for console execution? That way the >> dependencies could be kept relatively clean. > >It depends on either X11R6 or aout-svgalib - if any wants me to set it for >both, I'll consider it. I don't think separate .deb's are a good idea. The problem I had was I had not installed aout-sgvalib, and on running sabuse ldd complained about svgalib being the wrong exec format (it picked up the elf version instead). It took me a while to figure out what was wrong. Perhaps even pointing this out in /usr/doc/abuse would be a distinct improvement. Thanks again Gordon
Release of setserial-2.10-9
Package: setserial Version: 2.10-9 New release of the setserial package, fixing a number of outstanding bug reports. Note that the last release by me was 2.10-7, and it is on this version that the bug fixes were applied. -rw-rw-rw- 1 root root 5445 Sep 5 15:37 setserial-2.10-9.diff.gz -rw-rw-rw- 1 root root13109 Sep 5 15:37 setserial-2.10-9.tar.gz -rw-r--r-- 1 root root12402 Sep 5 15:37 setserial_2.10-9_i386.deb Gordon
Disk Performance
I am currently playing around with vmware, running win98. However, the
performance stinks (I am using a beta release though). The strange thing
is though that if I do a
find / -print > /dev/null
in another window, the performance IMPROVES...
Now I am worried that there is a bigger question here...
My / files are all on a SCSI disk, while vmware runs off an IDE drive.
Could this have something to do with interrupts? Or DMA settings?
Any thoughts?
My IDE drive is an UIDE drive, but I cannot find a way to tell the kernel.
Does it automatically know? Its a Quantum Fireball 10GByte. It seems
to know it is dma (using_dma is set).
Here is my /proc/ide/ide1/hdc/settings...
namevalue min max mode
- --- ---
bios_cyl19885 0 65535 rw
bios_head 16 0 255 rw
bios_sect 63 0 63 rw
breada_readahead4 0 127 rw
bswap 0 0 1 r
file_readahead 124 0 2097151 rw
io_32bit0 0 3 rw
keepsettings0 0 1 rw
max_kb_per_request 64 1 127 rw
multcount 0 0 8 rw
nice1 1 0 1 rw
nowerr 0 0 1 rw
pio_modewrite-only 0 255 w
slow0 0 1 rw
unmaskirq 0 0 1 rw
using_dma 1 0 1 rw
what is nice1? and pio_mode write-only looks curious...
Here is the output from hdparm -i /dev/hdc
/dev/hdc:
Model=QUANTUM FIREBALL CX10.2A, FwRev=A3F.0B00, SerialNo=833920130551
Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs }
RawCHS=19885/16/63, TrkSize=32256, SectSize=21298, ECCbytes=4
BuffType=3(DualPortCache), BuffSize=418kB, MaxMultSect=16, MultSect=off
DblWordIO=no, maxPIO=2(fast), DMA=yes, maxDMA=2(fast)
CurCHS=19885/16/63, CurSects=20044080, LBA=yes
LBA CHS=621/512/63 Remapping, LBA=yes, LBAsects=20044080
tDMA={min:120,rec:120}, DMA modes: mword0 mword1 *mword2
IORDY=on/off, tPIO={min:120,w/IORDY:120}, PIO modes: mode3 mode4
UDMA modes: mode0 mode1 mode2
Here is the output from hdparm /dev/hdc
/dev/hdc:
multcount= 0 (off)
I/O support = 0 (default 16-bit)
unmaskirq= 0 (off)
using_dma= 1 (on)
keepsettings = 0 (off)
nowerr = 0 (off)
readonly = 0 (off)
readahead= 8 (on)
geometry = 19885/16/63, sectors = 20044080, start = 0
I guess hdparm -c 1 -m 16 might improve things, but why does it go faster
when I use the scsi disk??
/proc/interrupts is
CPU0
0:6189709 XT-PIC timer
1: 11676 XT-PIC keyboard
2: 0 XT-PIC cascade
4: 89680 XT-PIC serial
5: 146029 XT-PIC aic7xxx
8:1539593 XT-PIC rtc
9: 265750 XT-PIC eth0
10: 1 XT-PIC soundblaster
13: 1 XT-PIC fpu
15: 44096 XT-PIC ide1
NMI: 0
ERR: 0
During bootup, the kernel (2.2.12) reports...
PIIX3: IDE controller on PCI bus 00 dev 39
PIIX3: not 100% native mode: will probe irqs later
ide0: BM-DMA at 0xffa0-0xffa7, BIOS settings: hda:pio, hdb:pio
ide1: BM-DMA at 0xffa8-0xffaf, BIOS settings: hdc:pio, hdd:pio
/dev/hdc is not detected by the bios. Incidentally, if I let the bios
detect the disk, the kernel starts hunting for /dev/hda, which causes a delay
of 10 seconds or more during bootup (I only have one IDE disk). Even then,
the BIOS never reports that it is a DMA disk (it can
detects the mode automatically only). Surely this delay indicates a problem
with the kernel?
All help appreciated. I have checked the HOWTOs, but they are seriously lacking
in useful information concerning this...
G.
--
Gordon Russell
http://www.dcs.napier.ac.uk/~gor
PGP Public Key - http://www.dcs.napier.ac.uk/~gor/pgpkey.txt
Problems with setserial and pcmcia
I am getting some reports that the new setserial 2.15-3 interferes with pcmcia cardservices. It would appear that setserial runs first, and initialises the pcmcia card for serial use. This is fine unless the card is a combo modem/ethernet card, in which case the ethernet part of the card is not initialised by cardservices. The bugno is 38131. Could someone more knowledgable than myself think about this one. I do not have a pcmcia setup, so it is clearly not a problem for me. I have changed both setserial code and the way it is started up. Previously setserial was an /etc/rc.boot, and now it is /etc/init.d script with rcS, rc0, and rc6 used. I used priority 30 in each directory, as I need modules (like serial) loaded before use, and need access to modules and /etc (I may do some writing) on shutdown and halts. Is this a problem with setserial or with pcmcia?? G. -- Gordon Russell http://www.dcs.napier.ac.uk/~gor PGP Public Key - http://www.dcs.napier.ac.uk/~gor/pgpkey.txt
