Re: recommended gcc compiler version
On 24 Dec 2000, Kai Henningsen wrote: > [EMAIL PROTECTED] (Anuradha Ratnaweera) wrote on 22.12.00 in ><[EMAIL PROTECTED]>: > > > On Fri, 22 Dec 2000, Alan Cox wrote: > > > > > For i386 > > > > > > 2.2.18 > > > gcc 2.7.2 or egcs-1.1.2 > > > > Just a remainder for debian users. There is a debian package gcc272 which > > is said to be the "GNU C compiler's C part", for "backword compatibility > > purposes". I recompiled my kernel after an > > > > apt-get install gcc272 > > > > and after setting > > > > HOSTGCC = gcc272 > > > > in kernel source tree Makerile. > > I recently compiled 2.2.18 and noticed that make-kpkg (from kernel-package > - don't compile kernels on Debian without it!) did that automatically. That is a very good thing. It would have been even better if the dependencies of the kernel-package does include gcc272 rather than giving a "command not found" error when make-kpkg is run without gcc272 installed. It might leave a new user clueless. Anuradha - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Abysmal RAID 0 performance on 2.4.0-test10 for IDE?
Felix von Leitner writes: > I bought 4 ATA-100 Maxtor drives and put them on a Promise Ultra100 > controller to make a single striping RAID of them to increase > throughput. > > I wrote a small test program that simply reads stdin linearly and > displays the throughput. Here are the results of my test program: > # rb < /dev/ide/host2/bus0/target0/lun0/part1 > 27.8 meg/sec > # rb < /dev/ide/host2/bus0/target0/lun0/part1 > 26.8 meg/sec > > Here is the result of my test program on the strip set: > # rb < /dev/md/0 > 30.3 meg/sec > hde: 160086528 sectors (81964 MB) w/2048KiB Cache, CHS=158816/16/63, UDMA(100) > hdf: 160086528 sectors (81964 MB) w/2048KiB Cache, CHS=158816/16/63, UDMA(100) > hdg: 160086528 sectors (81964 MB) w/2048KiB Cache, CHS=158816/16/63, UDMA(100) > hdh: 160086528 sectors (81964 MB) w/2048KiB Cache, CHS=158816/16/63, UDMA(100) That's because IDE doesn't allow multiple requests on the same bus, unlike SCSI. That's why IDE disks on the same bus are "master" and "slave". If you look at the 3ware IDE RAID systems, each drive has its own IDE bus. Maybe try a stripe set on only two disks, hde and hdg, and see how it works. Cheers, Andreas -- Andreas Dilger \ "If a man ate a pound of pasta and a pound of antipasto, \ would they cancel out, leaving him still hungry?" http://www-mddsp.enel.ucalgary.ca/People/adilger/ -- Dogbert - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Fatal Oops on boot with 2.4.0testX and recent GCC snapshots
On Mon, Dec 25, 2000 at 08:40:50PM + or thereabouts, Thorsten Kranzkowski wrote: > On Mon, Dec 25, 2000 at 06:09:35AM +0100, Mike Galbraith wrote: > > I wouldn't (not going to here;) spend a lot of time on it. The compiler > > has problems. It won't build glibc-2.2, and chokes horribly on ipchains. > > > > int ipt_register_table(struct ipt_table *table) > > { > > int ret; > > struct ipt_table_info *newinfo; > > static struct ipt_table_info bootstrap > > = { 0, 0, { 0 }, { 0 }, { } }; > >^ > > ip_tables.c:1361: Internal compiler error in array_size_for_constructor, at >varasm.c:4456 > > > Well, I 'fixed' this by changing above line to: > = { 0, 0, { 0 }, { 0 }, }; > and repeating this change (deleting the braces) about 15 times in 2 or 3 other > files of iptables. (patch available on request) > Of course gcc shouldn't die but issue a useful message if/when syntax rules > may have changed. > > Apart from that and a hand-edited arch/alpha/vmlinux.lds that got some > newlines wrong, the kernel compiled fine and is up for over a day now. > Though this is not intel but alpha (ev4 / AXPpci33). > > Marvin:~$ uname -a > Linux Marvin 2.4.0-test13pre4-ac2 #13 Sun Dec 24 15:26:57 UTC 2000 alpha unknown > Marvin:~$ uptime > 8:19pm up 1 day, 4:28, 4 users, load average: 0.00, 0.00, 0.00 > Marvin:~$ gcc -v > Reading specs from /usr/lib/gcc-lib/alpha-unknown-linux-gnu/2.97/specs > Configured with: ../gcc-20001211/configure --enable-threads --enable-shared >--prefix=/usr --enable-languages=c,c++ > gcc version 2.97 20001211 (experimental) > > > I use iptables for masquerading my local ethernet and that works as expected > so far. > > Thorsten. Its a problem with initializing a zero-length array. This is something that gcc has never previously been documented to do, but it has worked in the past (most of the time). Recently it has been decided (according to traffic on gcc-bugs and gcc-patches lists) that gcc will handle zero-length arrays as flexable-array-members per ISO C99 standard. AFAIK, that means that if they are to be initialized, zero-length arrays can only exist as the last element of a structure, and that the structure must not be embeded within another structure. The empty brackets that Thorsten removed were initializing the zero-length array to empty, but gcc currently has this bit of code in varasm.c (around line 4460): /* ??? I'm fairly certain if there were no elements, we shouldn't have created the constructor in the first place. */ if (max_index == NULL_TREE) abort (); This abort() resulted in the "Internal compiler error" that Mike noticed earlier. Removing the empty brackets prevents gcc from trying to initialize the zero length array and avoids this problem. However, this can result in warning messages about missing initializers depending upon the warning flags given to gcc, and seems like the wrong thing to do. The best solution (IMHO) for this situation is to change gcc/varasm.c to accept empty initializers, something like: /* ??? I'm fairly certain if there were no elements, we shouldn't have created the constructor in the first place. */ /* No, it can be useful to initialize the zero-length array with an empty initializer. */ if (max_index == NULL_TREE) return 0; The rest of netfilter will still not compile because in several other C files the initialized zero-length arrays are nested several structures deep. If we can convince the gcc folks to drop some of the ISO C99 restrictions on the use of zero-length arrays then all will be back to normal (as Ulrich Drepper pointed out, the ISO committee in their infinite wisdom does not always come up with a standard that is the best solution in the real world). But I am not sure if that is the best solution. Perhaps it would be better to change the netfilter code. In any event, the gcc documentation does not say anything about not being able to initialize zero-length arrays to empty, so this is a bug and I'm going to talk with the gcc folks. -Paul Laufer - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: innd mmap bug in 2.4.0-test12
On Tue, 26 Dec 2000, Chris Wedgwood wrote: > On Mon, Dec 25, 2000 at 01:42:33AM -0800, Linus Torvalds wrote: > > We just don't write them out. Because right now the only thing > that writes out dirty pages is memory pressure. "sync()", > "fsync()" and "fdatasync()" will happily ignore dirty pages > completely. The thing that made me overlook that simple thing in > testing was that I was testing the new VM stuff under heavy VM > load - to shake out any bugs. > > Does this mean anyone using test13-pre4 should also expect to see > data not being flushed on shutdown? No. This all only matters to things that do shared writable mmap's. Almost nothing does that. innd is (sadly) the only regular thing that uses this, which is why it's always innd that breaks, even if everything else works. And even innd is often compiled to use "write()" instead of shared mappings (it's a config option), so not even all innd's will break. Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
controllerless pci device support
would it be sensible to write a PCI device interface for controllerless PCI devices like serial PCI ports? I am now trying to make the older 2.2.x series LT winmodem patch into the 2.4.0-test13pre4 sources work. I see how some companies are unable to release all the source code to drivers due to legal reasons and patent restrictions. Maybe there should be a generic driver interface for software modems or other devices, so it is easier to - as an example - write winmodem drivers for the serial driver without hacking in many sets of "#ifdef LUCENT_MODEM. modified code #endif" to the serial.c source file. i am not able to create such a thing, and winmodems are not the most popular thing to talk about in regards to support. after spending 3 hours staring at serial.c, as a beginning programmer, and hand copying the appropriate 2.2.x winnmodem "serial.c" driver code in, i am lost. the module finally compiles, without error, but complains with an error that there is an unresolved symbol "jiffie". kind of funny, a jiffie is all that separates me from turning my brand new laptop into a machine i can use the modem on. also it is equally fustrating. will this situation improve in time or what else can i do to get my modem working? arrrgh! even if the hand-done patching of 2.4.x's serial.c file resulted in a useable kernel module, i would not like to have to patch it every time i update my kernel. a winmodem.o module with support for generic interfaces into the kernel so driver vendors do not need to muck around with serial.c would be an idea. my real question to all is where is the support of PCI serial devices at inside of the kernel? if i have pci bus 0:0.b sharing irq 11 with 0:0.c, does the linux kernel support both devices working at the same time (ethernet, and serial port aka winmodem)? this is probably better off sent to the serial mailing list i know, but i am more interested in whether all the problems i am having with 4 out of 6 devices on my laptop's PCI bus conflicting, whether this is because the linux kernel does not support more than one PCI function operating simultaneously on any given PCI device under the same PCI bus. ( bus:device.function ) right now i get a message that says [IRQ 11 is already used by device 0:8.0] when i load drivers for the device 0:8.1, and the visa-versa message when loading drivers for device 0:8.0. Is this just a warning, or an error? i can't tell. sometimes the driver (as is the case with pcmcia drivers, where slot0 is 0:6.0 and slot1 is 0:6.1) loads anyways, despite the message about [IRQ 11 is alr...]. othertimes, with my ethernet drivers and alsa sound drivers, i see the message and the drivers fail to load. what to do merry holidays, all. i apologize this is long and likely off topic. i mean well though. -Eric Shattow [EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: 8139too driver broken? (2.4-test12) - Was: Re: rtl8139 driver broken? (2.2.16)
: On Sat, 23 Dec 2000 18:50:53 +0100, Stefan Hoffmeister wrote: >The rather major problem that >remains is performance. In case someone is interested... Windows 2000 SP1 now has the Realtek 8139 (Celeron 433, 192 MB, pure SCSI); drivers as shipped with W2K. Using a 40 MB test file over FTP, I get Realtek card sends with 3.5 MB/s Realtek card receives with 5 MB/s The system that previously contained the 8139 card now has a (10 MBit) 8029 card - transfer rates with that card are about 850 KB/s, compared to the 400KB/s to 530 KB/s with the (100 MBit) 8139 card. This makes me conclude that there is some pretty serious problem left in the 8139too driver. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: innd mmap bug in 2.4.0-test12
On Mon, 25 Dec 2000, Marco d'Itri wrote: > > >Do you get any messages? I don't think you will, but it should be tested. > > I read you found the real cause so that may be bogus, but I have got two > messages while booting. The first showed up while doing the fsck of a 6 > GB file systems and killed the process (fscks of smaller partitions > completed successfully), the second occured while initializing > /dev/random and left an unkillable dd process and a stuck boot process > (I gathered this info with sysrq). I'd still love to get the trace for these. I think I have a handle on the problems, but it would stil be helpful - dropping a dirty page really shouldn't happen except for the swap cache (and that should have been plugged by adding the ClearPageDirty()). Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
RE: TCP keepalive seems to send to only one port
Cesar Barros wrote: > On Mon, Dec 25, 2000 at 04:33:07PM -0800, David Schwartz wrote: > > If the administrator of the NAT meant for you to have a > > permanent mapping, > > she would have put one there. Using keepalives to hold a NAT entry open > > indefinitely without activity would be considered abuse in most NAT > > configurations. The NAT might not consider a keepalive to be > > activity anyway > > (arguably, it shouldn't). > Well, consider the scenario of an application which opens a > control connection > and a data connection, and the data connection remains idle for some hours > while you get to the beginning of the queue, and then the > transfer starts. The > data connection is not open forever, and the timeout (and the > periodic pings) > is on the control connection. I would consider that application broken. The data connection should be opened when it's needed, not left idle for hours and used later. If the data connection does break somehow, there should be a provision for re-establishing it without losing all application-level state. I'm not saying it shouldn't be possible to work around a defective application protocol. But to expect there to be some easy way to just flip a switch and fix it is unreasonable. The NAT may not even consider a keepalive to be activity. There's no logical reason it should if the timeout is less than many hours. DS - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: TCP keepalive seems to send to only one port
On Mon, Dec 25, 2000 at 04:33:07PM -0800, David Schwartz wrote: > > > On Sat, Dec 23, 2000 at 04:19:31PM -0800, David Schwartz wrote: > > > > > This means that keepalive is useless for keeping alive more than > > > > one connection > > > > to a given host. > > > > Actually, keepalive is useless for keeping connections > > > alive anyway. It's > > > very badly named. It's purpose is to detect dead peers, not keep peers > > > alive. > > > > Then what do you do when you are behind a NAT? > > If the administrator of the NAT meant for you to have a permanent mapping, > she would have put one there. Using keepalives to hold a NAT entry open > indefinitely without activity would be considered abuse in most NAT > configurations. The NAT might not consider a keepalive to be activity anyway > (arguably, it shouldn't). Well, consider the scenario of an application which opens a control connection and a data connection, and the data connection remains idle for some hours while you get to the beginning of the queue, and then the transfer starts. The data connection is not open forever, and the timeout (and the periodic pings) is on the control connection. The problem is that, after four or more hours of waiting, when the other side finally starts sending, the NAT has already forgotten about the connection. (In case someone is wondering, the application I'm talking about is fidelio. Which is pretty useless on busy sites behind a NAT because of that.) -- Cesar Eduardo Barros [EMAIL PROTECTED] [EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: sparc 10 w/512 megs hangs during boot
apparently test7 has the same problem and when I compile test6 I get these errors: gcc -D__KERNEL__ -I/usr/src/linux/include -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -m32 -pipe -mno-fpu -fcall-used-g5 -fcall-used-g7 -fno-strict-aliasing-c -o fault.o fault.c gcc -D__KERNEL__ -I/usr/src/linux/include -Wall -Wstrict-prototypes -O2 -fomit-frame-pointer -m32 -pipe -mno-fpu -fcall-used-g5 -fcall-used-g7 -fno-strict-aliasing-c -o init.o init.c init.c: In function `kmap_init': init.c:92: warning: unused variable `pteval' init.c: In function `mem_init': init.c:460: `highmem_mapnr' undeclared (first use in this function) init.c:460: (Each undeclared identifier is reported only once init.c:460: for each function it appears in.) init.c: In function `flush_page_to_ram': init.c:588: warning: passing arg 1 of `___f___flush_page_to_ram' makes integer from pointer without a cast make[3]: *** [init.o] Error 1 make[3]: Leaving directory `/usr/src/linux/arch/sparc/mm' make[2]: *** [first_rule] Error 2 make[2]: Leaving directory `/usr/src/linux/arch/sparc/mm' make[1]: *** [_dir_arch/sparc/mm] Error 2 make[1]: Leaving directory `/usr/src/linux' make: *** [stamp-build] Error 2 so as far as I can see test5 is the last kernel that can be built and booted properly on a sparc10 with 512M of ram. All others after test5 do not boot my sparc10 with more than 128M of ram. ron --- Ron Calderon <[EMAIL PROTECTED]> wrote: > test8 is borked too. I'll try test7 next > > ron > --- Jan-Benedict Glaw <[EMAIL PROTECTED]> wrote: > > On Sun, Dec 24, 2000 at 12:48:44PM -0800, Ron > > Calderon wrote: > > > I just finished compiling 2.4.0-test5 and that > > worked > > > fine with 512M ram. I'll start going thru the > > other > > > kernels. It'll take me sometime since compileing > > takes > > > a long time. > > > > I've not yet started active searching. However: > > - test5 is fine > > - test13-pre3 is not > > > > I don't know how fast your machine is, but we > should > > coordinate out > > search... I'll try to build -test10final (with > > minimal config to > > only test boot) so that shouldn't take so very > > long... You should > > test sth around -test8... > > > > MfG, JBG > > > > -- > > Fehler eingestehen, Größe zeigen: Nehmt die > > Rechtschreibreform zurück!!! > > /* Jan-Benedict Glaw <[EMAIL PROTECTED]> -- > > +49-177-5601720 */ > > keyID=0x8399E1BB fingerprint=250D 3BCF 7127 0D8C > > A444 A961 1DBD 5E75 8399 E1BB > > "insmod vi.o and there we go..." (Alexander > > Viro on linux-kernel) > > > > > ATTACHMENT part 2 application/pgp-signature > > > > __ > Do You Yahoo!? > Yahoo! Shopping - Thousands of Stores. Millions of > Products. > http://shopping.yahoo.com/ > - > To unsubscribe from this list: send the line > "unsubscribe linux-kernel" in > the body of a message to [EMAIL PROTECTED] > Please read the FAQ at http://www.tux.org/lkml/ __ Do You Yahoo!? Yahoo! Shopping - Thousands of Stores. Millions of Products. http://shopping.yahoo.com/ - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: [bug] test13-pre4 nfs/ip_defrag crash (smp)
On Mon, 25 Dec 2000, Jasper Spaans wrote: > > I am having some reproducible crashes with 2.4.0-test13-pre4, whenever I > do some 'heavy' nfs-ing.. decoded oops: It looks like most of what you have is modules. Is netfilter enabled as a module too? Can you reproduce it without modules, in case it's a autounload race or similar? Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
RE: TCP keepalive seems to send to only one port
> On Sat, Dec 23, 2000 at 04:19:31PM -0800, David Schwartz wrote: > > > This means that keepalive is useless for keeping alive more than > > > one connection > > > to a given host. > > Actually, keepalive is useless for keeping connections > > alive anyway. It's > > very badly named. It's purpose is to detect dead peers, not keep peers > > alive. > > Then what do you do when you are behind a NAT? If the administrator of the NAT meant for you to have a permanent mapping, she would have put one there. Using keepalives to hold a NAT entry open indefinitely without activity would be considered abuse in most NAT configurations. The NAT might not consider a keepalive to be activity anyway (arguably, it shouldn't). > And how do you > expire entries in > ESTABLISHED state that could stay lingering forever without some sort of > keepalive? (The FINs might have been lost due to a conectivity > transient, so > you can have another perfectly valid and alive connection with > the same host, > and application-level timeouts are useless for some applications > (*cough*nc*cough*)) I don't understand this argument. First you demonstrate the need for application-level timeouts, then you say application-level timeouts are useless. Actually, your first argument is correct, protocols layered on top of TCP that don't provide for timeouts are defective. DS - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: About Celeron processor memory barrier problem
Tim Wright writes: > There was a similar thread to this recently. The issue is that if you > choose the wrong processor type, you may not even be able to complain. An illegal opcode handler could deal with the problem. It could crudely emulate just enough to make printk work. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH] LVM includes userlevel headers
On Tue, Dec 26, 2000 at 12:32:44AM +0100, Andrea Arcangeli wrote: > On Mon, Dec 25, 2000 at 11:53:33PM +0100, Christoph Hellwig wrote: > > The first patch fixes that and the second changes the toplevel Makefile > > to search only the kernel and gcc (for stdarg.h) includes to prevent such > > accidents. > > Looks fine, thanks. BTW, I included your fixes into the 2.2.x backport (nostdinc in a separate patch): ftp://ftp.us.kernel.org/pub/linux/kernel/people/andrea/kernels/v2.2/2.2.19pre3aa2/14_bigmem-rawio-lvm-0.9-2.2.19pre3aa2-3.bz2 ftp://ftp.us.kernel.org/pub/linux/kernel/people/andrea/kernels/v2.2/2.2.19pre3aa2/00_nostdinc-Christoph-Hellwig-1 Andrea - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Abysmal RAID 0 performance on 2.4.0-test10 for IDE?
Thus spake Felix von Leitner ([EMAIL PROTECTED]): > Here is the result of my test program on the strip set: > # rb < /dev/md/0 > 30.3 meg/sec > # One more detail: top says the CPU is 50% system when reading from either one of the disk or raid devices. That seems awfully high considering that the Promise controller claims to do UDMA. Any comments? Felix - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH] winbond-840 updates, tester needed!
Manfred wrote: > > I need a tester with a winbond card on a computer with a big endian cpu, > any volunteers? > Attached is a patch with further cleanups: * removed next_desc from tx descriptors - according to the documentation, it isn't needed. * limit each transmit buffer to < 1024 bytes. I still see several bad packets during ftp transfers, but that might be due to a bad cable. Noone out there with a ppc and a winbond card? -- Manfred --- 2.4/drivers/net/winbond-840.c Sun Dec 17 18:03:56 2000 +++ build-2.4/drivers/net/winbond-840.c Tue Dec 26 00:23:23 2000 @@ -21,11 +21,24 @@ Do not change the version information unless an improvement has been made. Merely removing my name, as Compex has done in the past, does not count as an improvement. + + Changelog: + * ported to 2.4 + ??? + * spin lock update, memory barriers, new style dma mappings + superfluous next pointers from tx ring removed, + limit tx buffer size to < 1024 + + Manfred Spraul <[EMAIL PROTECTED]> + + TODO: + * according to the documentation, the chip supports big endian + internally. Replace the cpu_to_le32 with that bit. */ /* These identify the driver base version and may not be removed. */ static const char version1[] = -"winbond-840.c:v1.01 5/15/2000 Donald Becker <[EMAIL PROTECTED]>\n"; +"winbond-840.c:v1.01 (2.4 port) 5/15/2000 Donald Becker <[EMAIL PROTECTED]>\n"; static const char version2[] = " http://www.scyld.com/network/drivers.html\n"; @@ -81,6 +94,8 @@ #define TX_FIFO_SIZE (2048) #define TX_BUG_FIFO_LIMIT (TX_FIFO_SIZE-1514-16) +#define TX_BUFLIMIT(1024-128) + /* Operational parameters that usually are not changed. */ /* Time in jiffies before concluding the transmitter is hung. */ #define TX_TIMEOUT (2*HZ) @@ -114,12 +129,6 @@ #include #include -/* Condensed operations for readability. - The compatibility defines are in kern_compat.h */ - -#define virt_to_le32desc(addr) cpu_to_le32(virt_to_bus(addr)) -#define le32desc_to_virt(addr) bus_to_virt(le32_to_cpu(addr)) - MODULE_AUTHOR("Donald Becker <[EMAIL PROTECTED]>"); MODULE_DESCRIPTION("Winbond W89c840 Ethernet driver"); MODULE_PARM(max_interrupt_work, "i"); @@ -280,7 +289,7 @@ s32 status; s32 length; u32 buffer1; - u32 next_desc; + u32 buffer2; }; struct w840_tx_desc { @@ -298,9 +307,12 @@ #define PRIV_ALIGN 15 /* Required alignment mask */ struct netdev_private { - /* Descriptor rings first for alignment. */ - struct w840_rx_desc rx_ring[RX_RING_SIZE]; - struct w840_tx_desc tx_ring[TX_RING_SIZE]; + struct w840_rx_desc *rx_ring; + dma_addr_t rx_addr[RX_RING_SIZE]; + struct w840_tx_desc *tx_ring; + dma_addr_t tx_addr[RX_RING_SIZE]; + dma_addr_t ring_dma_addr; + struct pci_dev *pdev; /* The addresses of receive-in-place skbuffs. */ struct sk_buff* rx_skbuff[RX_RING_SIZE]; /* The saved address of a sent-in-place packet/buffer, for later free(). */ @@ -335,7 +347,7 @@ static void check_duplex(struct net_device *dev); static void netdev_timer(unsigned long data); static void tx_timeout(struct net_device *dev); -static void init_ring(struct net_device *dev); +static int init_ring(struct net_device *dev); static int start_tx(struct sk_buff *skb, struct net_device *dev); static void intr_handler(int irq, void *dev_instance, struct pt_regs *regs); static void netdev_error(struct net_device *dev, int intr_status); @@ -364,6 +376,11 @@ return -EIO; pci_set_master(pdev); + if(!pci_dma_supported(pdev,0x)) { + printk(KERN_WARNING "Winbond-840: Device %s disabled due to DMA +limitations.\n", + pdev->name); + return -EIO; + } dev = init_etherdev(NULL, sizeof(*np)); if (!dev) return -ENOMEM; @@ -403,6 +420,7 @@ np = dev->priv; np->chip_id = chip_idx; np->drv_flags = pci_id_tbl[chip_idx].drv_flags; + np->pdev = pdev; spin_lock_init(>lock); pdev->driver_data = dev; @@ -632,10 +650,12 @@ printk(KERN_DEBUG "%s: w89c840_open() irq %d.\n", dev->name, dev->irq); - init_ring(dev); + if((i=init_ring(dev))) + return i; - writel(virt_to_bus(np->rx_ring), ioaddr + RxRingPtr); - writel(virt_to_bus(np->tx_ring), ioaddr + TxRingPtr); + writel(np->ring_dma_addr, ioaddr + RxRingPtr); + writel(np->ring_dma_addr+sizeof(struct w840_rx_desc)*RX_RING_SIZE, + ioaddr + TxRingPtr); for (i = 0; i < 6; i++) writeb(dev->dev_addr[i], ioaddr + StationAddr + i); @@ -733,11 +753,13 @@ "config %8.8x.\n", dev->name,
Re: [PATCH] LVM includes userlevel headers
On Mon, Dec 25, 2000 at 11:53:33PM +0100, Christoph Hellwig wrote: > The first patch fixes that and the second changes the toplevel Makefile > to search only the kernel and gcc (for stdarg.h) includes to prevent such > accidents. Looks fine, thanks. Andrea - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Abysmal RAID 0 performance on 2.4.0-test10 for IDE?
Hi, I bought 4 ATA-100 Maxtor drives and put them on a Promise Ultra100 controller to make a single striping RAID of them to increase throughput. I wrote a small test program that simply reads stdin linearly and displays the throughput. The block size is 100k. This is the result: # cat /etc/raidtab raiddev /dev/md/0 raid-level 0 nr-raid-disks 4 persistent-superblock 1 chunk-size 32 device /dev/ide/host2/bus0/target0/lun0/part1 raid-disk 0 device /dev/ide/host2/bus0/target1/lun0/part1 raid-disk 2 device /dev/ide/host2/bus1/target0/lun0/part1 raid-disk 1 device /dev/ide/host2/bus1/target1/lun0/part1 raid-disk 3 Here are the results of my test program on the disk devices: # rb < /dev/ide/host2/bus0/target0/lun0/part1 27.8 meg/sec # rb < /dev/ide/host2/bus0/target0/lun0/part1 26.8 meg/sec the other two disks have approximately the same numbers. Here is the result of my test program on the strip set: # rb < /dev/md/0 30.3 meg/sec # While this is faster than linear mode, I would have expected much better performance. These are the boot messages of the Promise adapter: PDC20267: IDE controller on PCI bus 00 dev 60 PDC20267: chipset revision 2 PDC20267: not 100% native mode: will probe irqs later PDC20267: (U)DMA Burst Bit ENABLED Primary PCI Mode Secondary PCI Mode. ide2: BM-DMA at 0xec00-0xec07, BIOS settings: hde:pio, hdf:pio ide3: BM-DMA at 0xec08-0xec0f, BIOS settings: hdg:pio, hdh:pio ide2 at 0xdc00-0xdc07,0xe002 on irq 10 ide3 at 0xe400-0xe407,0xe802 on irq 10 hde: 160086528 sectors (81964 MB) w/2048KiB Cache, CHS=158816/16/63, UDMA(100) hdf: 160086528 sectors (81964 MB) w/2048KiB Cache, CHS=158816/16/63, UDMA(100) hdg: 160086528 sectors (81964 MB) w/2048KiB Cache, CHS=158816/16/63, UDMA(100) hdh: 160086528 sectors (81964 MB) w/2048KiB Cache, CHS=158816/16/63, UDMA(100) I tuned the devices with hdparm -c 1 -a 32 -m 16 -p -u 1, for what it's worth (did not increase throughput but appeared to lessen the CPU usage). To verify that this is not an issue of the Promise controller, I started two instances of my test tool at the same time, one working on hde, the other on hdg (the two channels). Both yielded approximately 25 meg/sec, so it does not appear to be a hardware or driver issue. Is the RAID code really this slow? Any ideas what I can do? I am using the user space tools from raidtools-19990421-0.90.tar.bz2, but that should not have any influence, right? I heard that there is a new, faster RAID code somewhere, but it only claimed to be faster on RAID level 5, not on striping. Any tuning advice? By the way: I noticed another thing: one of the Maxtor hard disks was broken. It caused the whole box to freeze solid (no numlock, no console switches, no sysrq). That to me severely limits the usefulness of IDE RAID. While SCSI problems cause trouble, too, I have never seen one cause a complete freeze. How am I supposed to hot-swap the disks? I am using VESA framebuffer, so maybe there was a panic and it simply did not appear on my screen (or in the logs). Hope to hear from you soon (the RAID is needed on Dec 27). Should I use LVM instead of the MD code? Felix - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: [PATCH][RFC] LVM proc fix
On Mon, Dec 25, 2000 at 11:59:51PM +0100, Christoph Hellwig wrote: > Hi Linus & Heinz, > > there has been some discussion about the LVM /proc #ifdefs in > Linux 2.4.0-test13pre4 (LVM 0.9). How about just removing > CONFIG_LVM_PROC_FS? - beople that use LVM and procfs usually do > not care for the few extra bytes. I think it's a good idea. Andrea - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
[PATCH][RFC] LVM proc fix
Hi Linus & Heinz, there has been some discussion about the LVM /proc #ifdefs in Linux 2.4.0-test13pre4 (LVM 0.9). How about just removing CONFIG_LVM_PROC_FS? - beople that use LVM and procfs usually do not care for the few extra bytes. Patch attached. Christoph -- Whip me. Beat me. Make me maintain AIX. diff -uNr --exclude-from=dontdiff linux-2.4.0-test13-pre4/Documentation/Configure.help linux/Documentation/Configure.help --- linux-2.4.0-test13-pre4/Documentation/Configure.helpMon Dec 25 19:21:14 2000 +++ linux/Documentation/Configure.help Mon Dec 25 23:55:07 2000 @@ -1450,15 +1450,6 @@ want), say M here and read Documentation/modules.txt. The module will be called lvm-mod.o. -Logical Volume Manager /proc file system information -CONFIG_LVM_PROC_FS - If you say Y here, you are able to access overall Logical Volume - Manager, Volume Group, Logical and Physical Volume information in - /proc/lvm. - - To use this option, you have to check, that the "/proc file system - support" (CONFIG_PROC_FS) is enabled too. - Multiple devices driver support CONFIG_BLK_DEV_MD This driver lets you combine several hard disk partitions into one diff -uNr --exclude-from=dontdiff linux-2.4.0-test13-pre4/drivers/md/Config.in linux/drivers/md/Config.in --- linux-2.4.0-test13-pre4/drivers/md/Config.inSun Nov 26 17:23:18 2000 +++ linux/drivers/md/Config.in Mon Dec 25 23:55:07 2000 @@ -17,6 +17,5 @@ fi dep_tristate ' Logical volume manager (LVM) support' CONFIG_BLK_DEV_LVM $CONFIG_MD -dep_mbool ' LVM information in proc filesystem' CONFIG_LVM_PROC_FS $CONFIG_BLK_DEV_LVM endmenu diff -uNr --exclude-from=dontdiff linux-2.4.0-test13-pre4/drivers/md/lvm.c linux/drivers/md/lvm.c --- linux-2.4.0-test13-pre4/drivers/md/lvm.cMon Dec 25 19:21:16 2000 +++ linux/drivers/md/lvm.c Mon Dec 25 23:55:07 2000 @@ -139,6 +139,7 @@ * lvm_proc_get_global_info() *02/11/2000 - implemented /proc/lvm/ hierarchy *07/12/2000 - make sure lvm_make_request_fn returns correct value - 0 or 1 - NeilBrown + *25/12/2000 - fix procfs #defines - Christoph Hellwig * */ @@ -224,7 +225,7 @@ static int lvm_chr_ioctl(struct inode *, struct file *, uint, ulong); -#if defined CONFIG_LVM_PROC_FS && defined CONFIG_PROC_FS +#ifdef CONFIG_PROC_FS int lvm_proc_read_vg_info(char *, char **, off_t, int, int *, void *); int lvm_proc_read_lv_info(char *, char **, off_t, int, int *, void *); int lvm_proc_read_pv_info(char *, char **, off_t, int, int *, void *); @@ -347,7 +348,7 @@ static spinlock_t lvm_lock = SPIN_LOCK_UNLOCKED; static spinlock_t lvm_snapshot_lock = SPIN_LOCK_UNLOCKED; -#if defined CONFIG_LVM_PROC_FS && defined CONFIG_PROC_FS +#ifdef CONFIG_PROC_FS static struct proc_dir_entry *lvm_proc_dir = NULL; static struct proc_dir_entry *lvm_proc_vg_subdir = NULL; struct proc_dir_entry *pde = NULL; @@ -433,7 +434,7 @@ _chr_fops, NULL); #endif -#if defined CONFIG_LVM_PROC_FS && defined CONFIG_PROC_FS +#ifdef CONFIG_PROC_FS lvm_proc_dir = create_proc_entry (LVM_DIR, S_IFDIR, _root); if (lvm_proc_dir != NULL) { lvm_proc_vg_subdir = create_proc_entry (LVM_VG_SUBDIR, S_IFDIR, lvm_proc_dir); @@ -521,7 +522,7 @@ blksize_size[MAJOR_NR] = NULL; hardsect_size[MAJOR_NR] = NULL; -#if defined CONFIG_LVM_PROC_FS && defined CONFIG_PROC_FS +#ifdef CONFIG_PROC_FS remove_proc_entry(LVM_GLOBAL, lvm_proc_dir); remove_proc_entry(LVM_VG_SUBDIR, lvm_proc_dir); remove_proc_entry(LVM_DIR, _root); @@ -1263,7 +1264,7 @@ } -#if defined CONFIG_LVM_PROC_FS && defined CONFIG_PROC_FS +#ifdef CONFIG_PROC_FS /* * Support functions /proc-Filesystem */ @@ -1452,8 +1453,6 @@ else return count; } /* lvm_proc_get_global_info() */ -#endif /* #if defined CONFIG_LVM_PROC_FS && defined CONFIG_PROC_FS */ - /* * provide VG information @@ -1530,7 +1529,7 @@ return sz; } - +#endif /* CONFIG_PROC_FS */ /* * block device support function for /usr/src/linux/drivers/block/ll_rw_blk.c @@ -1989,7 +1988,7 @@ _chr_fops, NULL); #endif -#if defined CONFIG_LVM_PROC_FS && defined CONFIG_PROC_FS +#ifdef CONFIG_PROC_FS lvm_do_create_proc_entry_of_vg ( vg_ptr); #endif @@ -2021,7 +2020,9 @@ for (p = 0; p < vg_ptr->pv_max; p++) { if ( ( pv_ptr = vg_ptr->pv[p]) == NULL) { ret = lvm_do_pv_create(arg, vg_ptr, p); +#ifdef CONFIG_PROC_FS lvm_do_create_proc_entry_of_pv ( vg_ptr, pv_ptr); +#endif if ( ret != 0) return ret; /* We don't need the PE list @@ -2091,7 +2092,7 @@ if (copy_from_user(vg_name, arg, sizeof(vg_name)) != 0) return -EFAULT; -#if defined CONFIG_LVM_PROC_FS && defined CONFIG_PROC_FS +#ifdef CONFIG_PROC_FS
[PATCH] LVM includes userlevel headers
Hi Linus, LVM 0.9 that just rolled into 2.4-test includes userlevel headers - just to use constants from there to dublicate kernel functions. The first patch fixes that and the second changes the toplevel Makefile to search only the kernel and gcc (for stdarg.h) includes to prevent such accidents. Christoph -- Whip me. Beat me. Make me maintain AIX. diff -uNr --exclude-from=dontdiff linux-2.4.0-test13-pre4/drivers/md/lvm-snap.c linux/drivers/md/lvm-snap.c --- linux-2.4.0-test13-pre4/drivers/md/lvm-snap.c Mon Dec 25 19:21:16 2000 +++ linux/drivers/md/lvm-snap.c Mon Dec 25 23:59:50 2000 @@ -214,10 +214,10 @@ memset(lv_COW_table, 0, blksize_snap); for ( ; is < lv_snap->lv_remap_ptr; is++, id++) { /* store new COW_table entry */ - lv_COW_table[id].pv_org_number = LVM_TO_DISK64(lvm_pv_get_number(vg, lv_snap->lv_block_exception[is].rdev_org)); - lv_COW_table[id].pv_org_rsector = LVM_TO_DISK64(lv_snap->lv_block_exception[is].rsector_org); - lv_COW_table[id].pv_snap_number = LVM_TO_DISK64(lvm_pv_get_number(vg, lv_snap->lv_block_exception[is].rdev_new)); - lv_COW_table[id].pv_snap_rsector = LVM_TO_DISK64(lv_snap->lv_block_exception[is].rsector_new); + lv_COW_table[id].pv_org_number = cpu_to_le64(lvm_pv_get_number(vg, +lv_snap->lv_block_exception[is].rdev_org)); + lv_COW_table[id].pv_org_rsector = +cpu_to_le64(lv_snap->lv_block_exception[is].rsector_org); + lv_COW_table[id].pv_snap_number = cpu_to_le64(lvm_pv_get_number(vg, +lv_snap->lv_block_exception[is].rdev_new)); + lv_COW_table[id].pv_snap_rsector = +cpu_to_le64(lv_snap->lv_block_exception[is].rsector_new); } } @@ -268,10 +268,10 @@ blocks[0] = (snap_pe_start + COW_table_sector_offset) >> (blksize_snap >> 10); /* store new COW_table entry */ - lv_COW_table[idx_COW_table].pv_org_number = LVM_TO_DISK64(lvm_pv_get_number(vg, lv_snap->lv_block_exception[idx].rdev_org)); - lv_COW_table[idx_COW_table].pv_org_rsector = LVM_TO_DISK64(lv_snap->lv_block_exception[idx].rsector_org); - lv_COW_table[idx_COW_table].pv_snap_number = LVM_TO_DISK64(lvm_pv_get_number(vg, snap_phys_dev)); - lv_COW_table[idx_COW_table].pv_snap_rsector = LVM_TO_DISK64(lv_snap->lv_block_exception[idx].rsector_new); + lv_COW_table[idx_COW_table].pv_org_number = cpu_to_le64(lvm_pv_get_number(vg, +lv_snap->lv_block_exception[idx].rdev_org)); + lv_COW_table[idx_COW_table].pv_org_rsector = +cpu_to_le64(lv_snap->lv_block_exception[idx].rsector_org); + lv_COW_table[idx_COW_table].pv_snap_number = cpu_to_le64(lvm_pv_get_number(vg, +snap_phys_dev)); + lv_COW_table[idx_COW_table].pv_snap_rsector = +cpu_to_le64(lv_snap->lv_block_exception[idx].rsector_new); length_tmp = iobuf->length; iobuf->length = blksize_snap; diff -uNr --exclude-from=dontdiff linux-2.4.0-test13-pre4/include/linux/lvm.h linux/include/linux/lvm.h --- linux-2.4.0-test13-pre4/include/linux/lvm.h Mon Dec 25 19:21:15 2000 +++ linux/include/linux/lvm.h Tue Dec 26 00:01:23 2000 @@ -57,6 +57,8 @@ *26/06/2000 - implemented snapshot persistency and resizing support *02/11/2000 - added hash table size member to lv structure *12/11/2000 - removed unneeded timestamp definitions + *24/12/2000 - removed LVM_TO_{CORE,DISK}*, use cpu_{from, to}_le* + * instead - Christoph Hellwig * */ @@ -67,7 +69,6 @@ #define_LVM_KERNEL_H_VERSION "LVM 0.9 (13/11/2000)" #include -#include /* * preprocessor definitions @@ -323,51 +324,6 @@ COW_table_entries_per_PE = LVM_GET_COW_TABLE_CHUNKS_PER_PE(vg, lv); \ COW_table_chunks_per_PE = ( COW_table_entries_per_PE * sizeof(lv_COW_table_disk_t) / SECTOR_SIZE + lv->lv_chunk_size - 1) / lv->lv_chunk_size; \ COW_table_entries_per_PE - COW_table_chunks_per_PE;}) - - -/* to disk and to core data conversion macros */ -#if __BYTE_ORDER == __BIG_ENDIAN - -#define LVM_TO_CORE16(x) ( \ -((uint16_t)uint16_t)(x) & 0x00FFU) << 8) | \ -(((uint16_t)(x) & 0xFF00U) >> 8 - -#define LVM_TO_DISK16(x) LVM_TO_CORE16(x) - -#define LVM_TO_CORE32(x) ( \ -((uint32_t)uint32_t)(x) & 0x00FFU) << 24) | \ -(((uint32_t)(x) & 0xFF00U) << 8))) \ -(((uint32_t)(x) & 0x00FFU) >> 8))) \ -(((uint32_t)(x) & 0xFF00U) >> 24 - -#define LVM_TO_DISK32(x) LVM_TO_CORE32(x) - -#define LVM_TO_CORE64(x) \ -((uint64_t)uint64_t)(x) & 0x00FFULL) << 56) | \ -(((uint64_t)(x) & 0xFF00ULL) << 40) | \ -(((uint64_t)(x) & 0x00FFULL) << 24) | \ -(((uint64_t)(x) & 0xFF00ULL) << 8) | \ -(((uint64_t)(x) & 0x00FFULL) >> 8) | \ -
[bug] test13-pre4 nfs/ip_defrag crash (smp)
Hi! I am having some reproducible crashes with 2.4.0-test13-pre4, whenever I do some 'heavy' nfs-ing.. decoded oops: ksymoops 2.3.4 on i686 2.4.0-test13-pre4. Options used -V (default) -k /proc/ksyms (default) -l /proc/modules (default) -o /lib/modules/2.4.0-test13-pre4/ (default) -m /boot/System.map-2.4.0-test13-pre4 (default) Warning: You did not tell me where to find symbol information. I will assume that the log matches the kernel and modules that are running right now and I'll use the default options above for symbol resolution. If the current kernel and/or modules do not match the log, you can get more accurate output by telling me the kernel version and where to find map, modules, ksyms etc. ksymoops -h explains the options. Unable to handle kernel paging request at virtual address eabc089f c01e263e *pde = Oops: CPU:1 EIP:0010:[] Using defaults from ksymoops -t elf32-i386 -a i386 EFLAGS: 00010286 eax: eabc0863 ebx: c40c53e0 ecx: c54a1824 edx: c28079a0 esi: 0b90 edi: c5a0df40 ebp: 05c8 esp: ce699c4c ds: 0018 es: 0018 ss: 0018 Process rpciod (pid: 2358, stackpage=ce699000) Stack: c28079a0 0a2b 070a 0014 c01e2a2d c28079a0 c5a0df40 c02ce7ac ce698000 c36f9c40 c5a0df40 eabc0863 c54a1810 d1147a82 c5a0df40 ce699d48 c02f6878 c01e549c ce699d58 d11471c9 c5a0df40 ce699d48 Call Trace: [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] Code: 8b 40 3c 8b 4c 24 1c 89 41 3c c7 47 18 00 00 00 00 8b 54 24 >>EIP; c01e263e<= Trace; c01e2a2d Trace; eabc0863 Trace; d1147a82 <[8139too]rtl8139_set_rx_mode+56/270> Trace; c01e549c Trace; d11471c9 <[8139too]rtl8139_rx_interrupt+151/24c> Trace; c01e549c Trace; c012f810 <__alloc_pages+12c/2d4> Trace; d114640a <[8139too]rtl8139_hw_start+226/574> Trace; c01e549c Trace; c01dc838 Trace; c01e549c Trace; c01e549c Trace; c01dcab7 Trace; c01e549c Trace; d114913c <[8139too].data.end+51d/43e1> Trace; c01e4a3b Trace; c01e549c Trace; c01fa5f8 Trace; c01d71d5 Trace; c01e4b66 Trace; c01fa5f8 Trace; ea0a Trace; c01d50ae <__kfree_skb+132/138> Trace; c01faa8e Trace; c01fa5f8 Trace; ea0a Trace; ea0a Trace; ea0a Trace; c0200236 Trace; c01d21e5 Trace; d110b84a <[uhci]uhci_show_queues+13e/228> Trace; c01e1cd6 Trace; c01e1e0c Trace; d11677d0 <[ipt_LOG].data.end+3bd9/b469> Trace; d1170ee8 <.bss.end+1465/> Trace; d110e7e9 <[uhci]alloc_uhci+2c1/310> Trace; d110b6f5 <[uhci]uhci_is_skeleton_qh+19/30> Trace; d11099f3 <[usbcore]usbdevfs_root_inode_operations+13/40> Trace; d110cf3b <[uhci]uhci_submit_bulk+1b3/238> Trace; d110d264 <[uhci]uhci_result_isochronous+4/a4> Trace; d110d339 <[uhci]uhci_find_urb_ep+35/d0> Trace; d110dc2b <[uhci]rh_submit_urb+7b/670> Trace; d1116cc4 <[sunrpc]xprt_reserve_status+74/7c> Trace; d1116cc4 <[sunrpc]xprt_reserve_status+74/7c> Trace; d1116cbc <[sunrpc]xprt_reserve_status+6c/7c> Trace; d1116cbc <[sunrpc]xprt_reserve_status+6c/7c> Trace; c0107480 Trace; d1116cc4 <[sunrpc]xprt_reserve_status+74/7c> Trace; d1116cd0 <[sunrpc]xprt_request_init+4/90> Code; c01e263e <_EIP>: Code; c01e263e<= 0: 8b 40 3c mov0x3c(%eax),%eax <= Code; c01e2641 3: 8b 4c 24 1c mov0x1c(%esp,1),%ecx Code; c01e2645 7: 89 41 3c mov%eax,0x3c(%ecx) Code; c01e2648 a: c7 47 18 00 00 00 00 movl $0x0,0x18(%edi) Code; c01e264f 11: 8b 54 24 00 mov0x0(%esp,1),%edx Kernel panic: Aiee, killing interrupt handler! Regards, -- Jasper Spaans <[EMAIL PROTECTED]> - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
2.4-test5 mkisofs corruption
I have not yet tested later versions, but 2.4-test5 corrupts my jpegs. I made an iso image and mounted it in loopback and they are corrupted; it is not loopback doing it, I burned a CD and they were corrupted the same way. I downgraded to 2.2.18 and that works fine. If this is a known problem and/or has been addressed, great. If not, please ping me and I'll try the latest 2.4 test and see if the problem has gone away. Config: 900Mhz K7 on ASUS A7V MB PC100 mem w/ ECC (I don't think the board supports that; it's not enabled) IDE devices: /dev/hda is a Maxtor 91303D6, 12427MB w/512kB Cache, CHS=25249/16/63 /dev/hdb is a ATAPI 40X CD-ROM drive, 128kB Cache /dev/hdc is a IBM-DTLA-307045, 43979MB w/1916kB Cache, CHS=5606/255/63 /dev/hdd is a IBM-DTLA-307045, 43979MB w/1916kB Cache, CHS=5606/255/63 SCSI devices: /dev/sr0 is a MATSHITA CD-ROM, model CD-R CW-7502 4 ethernet interfaces eth0: 3Com 3c905B Cyclone 100baseTx eth1: 3Com 3c905B Cyclone 100baseTx eth2: 3Com 3c905B Cyclone 100baseTx eth3: 3Com 3c905 Boomerang 100baseTx PCI bus devices: Host bridge: VIA Technologies Unknown device (rev 2). PCI bridge: VIA Technologies Unknown device (rev 0). ISA bridge: VIA Technologies Unknown device (rev 34). IDE interface: VIA Technologies VT 82C586 Apollo IDE (rev 16). Host bridge: VIA Technologies Unknown device (rev 48). Ethernet controller: 3Com 3C905B 100bTX (rev 48). Ethernet controller: 3Com 3C905B 100bTX (rev 48). Ethernet controller: 3Com 3C905B 100bTX (rev 48). SCSI storage controller: Adaptec AIC-7850 (rev 3). Ethernet controller: 3Com 3C905 100bTX (rev 0). Unknown mass storage controller: Promise Technology Unknown device (rev 2). VGA compatible controller: Matrox Matrox G200 AGP (rev 1). -- --- Larry McVoy lm at bitmover.com http://www.bitmover.com/lm - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Fatal Oops on boot with 2.4.0testX and recent GCC snapshots
On Mon, Dec 25, 2000 at 06:09:35AM +0100, Mike Galbraith wrote: > I wouldn't (not going to here;) spend a lot of time on it. The compiler > has problems. It won't build glibc-2.2, and chokes horribly on ipchains. > > int ipt_register_table(struct ipt_table *table) > { > int ret; > struct ipt_table_info *newinfo; > static struct ipt_table_info bootstrap > = { 0, 0, { 0 }, { 0 }, { } }; >^ > ip_tables.c:1361: Internal compiler error in array_size_for_constructor, at >varasm.c:4456 Well, I 'fixed' this by changing above line to: = { 0, 0, { 0 }, { 0 }, }; and repeating this change (deleting the braces) about 15 times in 2 or 3 other files of iptables. (patch available on request) Of course gcc shouldn't die but issue a useful message if/when syntax rules may have changed. Apart from that and a hand-edited arch/alpha/vmlinux.lds that got some newlines wrong, the kernel compiled fine and is up for over a day now. Though this is not intel but alpha (ev4 / AXPpci33). Marvin:~$ uname -a Linux Marvin 2.4.0-test13pre4-ac2 #13 Sun Dec 24 15:26:57 UTC 2000 alpha unknown Marvin:~$ uptime 8:19pm up 1 day, 4:28, 4 users, load average: 0.00, 0.00, 0.00 Marvin:~$ gcc -v Reading specs from /usr/lib/gcc-lib/alpha-unknown-linux-gnu/2.97/specs Configured with: ../gcc-20001211/configure --enable-threads --enable-shared --prefix=/usr --enable-languages=c,c++ gcc version 2.97 20001211 (experimental) I use iptables for masquerading my local ethernet and that works as expected so far. Thorsten. -- | Thorsten KranzkowskiInternet: [EMAIL PROTECTED]| | Mobile: ++49 170 1876134 Snail: Niemannsweg 30, 49201 Dissen, Germany | | Ampr: dl8bcu@db0lj.#rpl.deu.eu, [EMAIL PROTECTED] [44.130.8.19] | - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: About Celeron processor memory barrier problem
> > [EMAIL PROTECTED] (Tim Wright) wrote on 24.12.00 in ><[EMAIL PROTECTED]>: > > > On Sun, Dec 24, 2000 at 11:36:00AM +0200, Kai Henningsen wrote: > > > There was a similar thread to this recently. The issue is that if you > > choose the wrong processor type, you may not even be able to complain. > > Hmm ... I think I can see ways around that (essentially similar to the 16 > bit bootstrap code), but it may indeed be more trouble than it's worth. What about a simple solution, "Ok, Booting the kernel for i486+fpu and above." - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: innd mmap bug in 2.4.0-test12
On Dec 25, Linus Torvalds <[EMAIL PROTECTED]> wrote: >Add a printk() to __remove_inode_page() that complains whenever it removes >a dirty page. > >Oh, in order to not see this with swap pages (which _can_ be removed when >they are dirty, if all users of them are gone), add a PageClearDirty() to >"remove_from_swap_cache()" so that we don't get false positives.. > >Do you get any messages? I don't think you will, but it should be tested. I read you found the real cause so that may be bogus, but I have got two messages while booting. The first showed up while doing the fsck of a 6 GB file systems and killed the process (fscks of smaller partitions completed successfully), the second occured while initializing /dev/random and left an unkillable dd process and a stuck boot process (I gathered this info with sysrq). Being -test12 unstable for me, if you don't need more data I'll go back to -test9 until the next release. >That's probably the infinite loop in the tty task queue handling, should >be fixed in test13-pre3 or so. Looks like I missed it, evil vger postmasters unsubscribed me again for no apparent reason... -- ciao, Marco - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: BIOS problem, pro Microsoft, anti other OS
Jacques > Can you check what does Linux 2.4.0-test behave, please? Are you comfortable compiling kernels? If not I'll compile one for you for the test. Currently I'm happily under 2.4.0-test 12: lsmod: Module Size Used by ppp_deflate39164 1 (autoclean) bsd_comp4148 0 (autoclean) ppp_async 6220 1 (autoclean) ppp_generic12820 2 (autoclean) [ppp_deflate bsd_comp ppp_async] ltmodem 364948 1 input 3328 0 serial 42192 1 (autoclean) (ONLY suporting my mouse) isa-pnp27528 0 (autoclean) [ltmodem serial] usbcore27684 0 (unused) Martin Are there any particular kernel config choices that will be beneficial for this problem? Mark > You probably should make the ltmodem driver check the region base > registers and interrupts and if they are not set, recommend the user to > change the OS or PNP settings in their BIOS setup. Can these be included in the LTmodem packages you are working up? MarvS - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Fatal Oops on boot with 2.4.0testX and recent GCC snapshots
On Mon, 25 Dec 2000, Andreas Franck wrote: > Hello Mike, hello linux-kernel hackers, > > Mike Galbraith wrote: > > I wouldn't (not going to here;) spend a lot of time on it. The compiler > > has problems. It won't build glibc-2.2, and chokes horribly on ipchains. > > Maybe, but you were lucky getting an ICE, and not silently failing code :-) You bet. > After having spent several hours debugging now, I think it was > worth it (at least for my understanding of lower-level kernel issues and of > the (rather nice and almost readable) assembly code gcc generates). There Don't get me wrong, chasing things like this is never a waste of time. In the case of gcc in particular. Our next 'stable' kernel compiler is going to come from the gcc development tree just as the next 'stable' kernel is coming out of the kernel development tree. -Mike - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: css hang; somewhere between test12 and test13pre4ac2
Hi Jens, Thanks for the patch; unfortunatly at the moment I can neither confirm or deny whether the patch works - since the damn thing won't crash on me with or without the patch - I suspect it depends on system mood. But I have a PDA logging console so if it happens I should catch it. Thanks again, Dave -- Have a happy GNU millennium! -- / Dr. David Alan Gilbert | Running GNU/Linux on | Happy \ \ gro.gilbert @ treblig.org | Alpha, x86, ARM and SPARC | In Hex / \ ___|___ http://www.treblig.org |/ - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Fatal Oops on boot with 2.4.0testX and recent GCC snapshots
On Mon, 25 Dec 2000, Andreas Franck wrote: > Hello Mike, hello linux-kernel hackers, > > Mike Galbraith wrote: > > I wouldn't (not going to here;) spend a lot of time on it. The compiler > > has problems. It won't build glibc-2.2, and chokes horribly on ipchains. > > Maybe, but after having spent several hours debugging now, I think it was > worth it: I am almost sure this is not a gcc bug, but a nasty race condition > involving the semaphore handling bdflush_init. > > I figured out by spilling some printk's around in bdflush_init, which made > the bug magically disappear, what wasn't what I intended - but which gave me > a clearer impression of what's going on. Oh? Can you show me (offline) what you did exactly that made it go away? (that's kinda scary.. _much_ prefer 'compiler has rough edges' option;) -Mike - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Netgear FA311
On Mon, 25 Dec 2000, Manfred wrote: > Could you try this setup? > > /* Configure the PCI bus bursts and FIFO thresholds. */ > /* Configure for standard, in-spec Ethernet. */ > np->tx_config = (1<<28) + /* Automatic transmit padding */ > (1<<23) + /* Excessive collision retry */ > (6<<20) + /* Max DMA burst = 128 byte */ > (8<<8) +/* fill threshold = 256 byte */ > 8; /* drain threshold = 256 byte */ > writel(np->tx_config, ioaddr + TxConfig); > Hmm, that little change worked a lot better. However thoughput is down to 700kb/s! Transferring files from the other machine to this machine is much faster - 868kb/s. In the logs, I only got *one* message from the natsemi driver (and this happened when sending files from this machine to the other machine. No problems receiving from the other machine. Dec 25 17:28:12 tahallah kernel: eth0: Something Wicked happened! 0583. But I just realised that the other machine I'm using has an 10 megabit ethernet card (on the hub that one is shown on 10, my machine is shown as 100). I think this explains the throughput problem. When I get another 4 way power socket, I'll put my other machine (which has a 100 megabit card in it) on the network and see if that makes a difference. Cheers, Alex -- Huffapuff! http://www.tahallah.clara.co.uk - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Netgear FA311
Alex wrote: > In the logs I'm seeing this: > > Dec 25 15:25:18 tahallah last message repeated 2 times > Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0783. > Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0780. 783 means: Tx Underrun Tx Idle Tx Packet Error Tx Descriptor Rx Packet Error Rx Descriptor Rx OK. Hmm. I download the Documentation from National (http://www.national.com/pf/DP/DP83815.html), and the the tx burst size/fill threshold/drain threshold combination is invalid: from natsemi.c: /* Configure the PCI bus bursts and FIFO thresholds. */ /* Configure for standard, in-spec Ethernet. */ np->tx_config = (1<<28) + /* Automatic transmit padding */ (1<<23) + /* Excessive collision retry */ (0x0<<20) + /* Max DMA burst = 512 byte */ (8<<8) +/* fill threshold = 256 byte */ 2; /* drain threshold = 64 byte */ writel(np->tx_config, ioaddr + TxConfig); But: page 51 The MXDMA MUST NOT be greater than the Tx Fill Threshold >> Could you try this setup? /* Configure the PCI bus bursts and FIFO thresholds. */ /* Configure for standard, in-spec Ethernet. */ np->tx_config = (1<<28) + /* Automatic transmit padding */ (1<<23) + /* Excessive collision retry */ (6<<20) + /* Max DMA burst = 128 byte */ (8<<8) +/* fill threshold = 256 byte */ 8; /* drain threshold = 256 byte */ writel(np->tx_config, ioaddr + TxConfig); -- Manfred - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: test13-pre4... udf problem with dvd access vs test12
On Mon, Dec 25 2000, Mohammad A. Haque wrote: > Fix confirmed. Am i supposed to get some DriveSense errors? I probably > am just don't recall. Good. The whole idea of cdrom_log_sense is to be able to cleanly limit the (often) sense-less (yes, haha :) ide-cd verbosity. Some of the css stuff are obvious candidates. We want users to be aware of the error, but often we end up flooding the logs with the same stuff over and over again. This is often a source of confusion for the casual user. -- * Jens Axboe <[EMAIL PROTECTED]> * SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
serial mouse - lockup connection 2.4.0-t12
greetings! the lockup of test12 leaves no droppings i can find, but in the course of a half-dozen lockups in the last few days i've made an observation or two that may be of diagnostic help. system is a k6-2 on a via chipset mb with onboard everything but video (fic va-503a). kernel built with gcc-2.95.2. glibc is 2.2. the lockups occur during mouse movement -- mouse is kensington emulating microsoft. the peculiar thing is that the mouse acceleration seems to slow a little in the seconds before the lockup. i don't know that this is a useful datum, but it seemed as if it might be significant. the lockups are otherwise neither predictable or reproducible. -- dep -- bipartisanship: an illogical construct not unlike the idea that if half the people like red and half the people like blue, the country's favorite color is purple. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: test13-pre4... udf problem with dvd access vs test12
Sorry, I meant DriveReady SeekComplete Error "Mohammad A. Haque" wrote: > Fix confirmed. Am i supposed to get some DriveSense errors? I probably > am just don't recall. -- = Mohammad A. Haque http://www.haque.net/ [EMAIL PROTECTED] "Alcohol and calculus don't mix. Project Lead Don't drink and derive." --Unknown http://wm.themes.org/ [EMAIL PROTECTED] = - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: test13-pre4... udf problem with dvd access vs test12
Fix confirmed. Am i supposed to get some DriveSense errors? I probably am just don't recall. Jens Axboe wrote: > Yes I know about this one, I've attached the patch here again. Linus, > could you apply? > > --- drivers/ide/ide-cd.c~ Sat Dec 23 23:59:52 2000 > +++ drivers/ide/ide-cd.cSun Dec 24 00:03:38 2000 > @@ -333,7 +333,7 @@ > { > int log = 0; > > - if (sense == NULL || pc->quiet) > + if (sense == NULL || pc == NULL || pc->quiet) > return 0; > > switch (sense->sense_key) { > > -- > * Jens Axboe <[EMAIL PROTECTED]> > * SuSE Labs -- = Mohammad A. Haque http://www.haque.net/ [EMAIL PROTECTED] "Alcohol and calculus don't mix. Project Lead Don't drink and derive." --Unknown http://wm.themes.org/ [EMAIL PROTECTED] = - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: test13-pre4... udf problem with dvd access vs test12
On Mon, Dec 25 2000, Mohammad A. Haque wrote: > Jens, I made sure to reverse the udf patch I mentioned in another thread > (all it really is merging changes from linux-udf cvs into the current > kernel). So this is from a clean test13-pre4 w/ some netfilter fixes. Ok, looks unrelated however. > >>EIP; c019c017<= Yes I know about this one, I've attached the patch here again. Linus, could you apply? --- drivers/ide/ide-cd.c~ Sat Dec 23 23:59:52 2000 +++ drivers/ide/ide-cd.cSun Dec 24 00:03:38 2000 @@ -333,7 +333,7 @@ { int log = 0; - if (sense == NULL || pc->quiet) + if (sense == NULL || pc == NULL || pc->quiet) return 0; switch (sense->sense_key) { -- * Jens Axboe <[EMAIL PROTECTED]> * SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Fatal Oops on boot with 2.4.0testX and recent GCC snapshots
Hello Mike, hello linux-kernel hackers, Mike Galbraith wrote: > I wouldn't (not going to here;) spend a lot of time on it. The compiler > has problems. It won't build glibc-2.2, and chokes horribly on ipchains. Maybe, but after having spent several hours debugging now, I think it was worth it: I am almost sure this is not a gcc bug, but a nasty race condition involving the semaphore handling bdflush_init. I figured out by spilling some printk's around in bdflush_init, which made the bug magically disappear, what wasn't what I intended - but which gave me a clearer impression of what's going on. It seems that whyever, the cause for this failure is actually the down(sem) call on a not yet up()'ed semaphore, and this is where it starts to get ugly. -- ->>>--- Andreas Franck <<<- ---<<< [EMAIL PROTECTED] --->>>--- ->>> Keep smiling! <<<- - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: BIOS problem, pro Microsoft, anti other OS
Hello! > This alert should probably be forwarded to Others, but appropriate > subTask persons in the kernel-source Maintainers list were not obvious. > > Briefly, documented below is the fact/complications that some PC BIOS > chips are now coming with a default Microsoft setting, which makes them > hostile to some functionalities of other OS. If particular under Linux, > a PCI Winmodem did NOT function with the Win98 BIOS setting, but did > fine with BIOS choice "Other OS". Possible, other PCI devices under > Linux OS might be simmilarly afflicated. > > This indicates a need for Linux install software to be equipped with a > utility to probe the BIOS and report back "Linux hostile" BIOS > settings. Today most Newbies are getting new PC boxes equipped with > WinModems. Hostile BIOS settings will block their capability to get > on-line. Unfortunately, I do not have the technical capablity to > directly contribute. Thus please forward this alert to however may be > capable and concerned with dealing with the problem. Can you check what does Linux 2.4.0-test behave, please? I know of these problems and I hope the new PCI code in 2.4.0 is able to assign the missing memory/IO resources without help of the BIOS, but unfortunately 2.2 isn't and it's very difficult to back-port the fixes as they depend on changes in many other parts of the kernel. You probably should make the ltmodem driver check the region base registers and interrupts and if they are not set, recommend the user to change the OS or PNP settings in their BIOS setup. Have a nice fortnight -- Martin `MJ' Mares <[EMAIL PROTECTED]> <[EMAIL PROTECTED]> http://atrey.karlin.mff.cuni.cz/~mj/ First law of socio-genetics: Celibacy is not hereditary. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Fatal Oops on boot with 2.4.0testX and recent GCC snapshots
Hello Mike, hello linux-kernel hackers, Mike Galbraith wrote: > I wouldn't (not going to here;) spend a lot of time on it. The compiler > has problems. It won't build glibc-2.2, and chokes horribly on ipchains. Maybe, but you were lucky getting an ICE, and not silently failing code :-) After having spent several hours debugging now, I think it was worth it (at least for my understanding of lower-level kernel issues and of the (rather nice and almost readable) assembly code gcc generates). There seems to be something going wrong in the down(sem) path after the kernel_thread call. I'm not sure if down() succeeds instantly when compiling the kernel with 2.95.2, but it seems to fail for 2.97; I figured out by spilling some printk's around in bdflush_init, which made the bug magically disappear, due to the looser timing. This also might happen for compiling with frame pointers or with the static declaration variables, somehow. Th bdflush_init function itself does not seem to be responsible, which corresponds with the assembly, which is fine and should get the same results for all compiled cases. It seems that whyever, the cause for this failure is actually the down(sem) call on a not yet up()'ed semaphore, and this is where it starts to get ugly. down() then calls __down_failed, which ends up in __down(); __down does some waitqueue handling, which I don't understand, and then calls __wake_up - up to then, everything seems fine, in __wake_up it is where my search ended up to now, but I think something is wrong in this context; however, the complexity of this code exceeds my knowledge by magnitudes, so I can't continue searching there without going mad :-) It would be nice if someone else could look from there on, now I've narrowed the case down to rather low-level functions. Greetings, Andreas -- ->>>--- Andreas Franck <<<- ---<<< [EMAIL PROTECTED] --->>>--- ->>> Keep smiling! <<<- - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: css hang; somewhere between test12 and test13pre4ac2
Heh, just go it. Testing Jens Axboe wrote: > > Thanks. If it's the cdrom_log_sense oops, please try also with > previously sent patch. -- = Mohammad A. Haque http://www.haque.net/ [EMAIL PROTECTED] "Alcohol and calculus don't mix. Project Lead Don't drink and derive." --Unknown http://wm.themes.org/ [EMAIL PROTECTED] = - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: test13-pre4... udf problem with dvd access vs test12
Jens, I made sure to reverse the udf patch I mentioned in another thread (all it really is merging changes from linux-udf cvs into the current kernel). So this is from a clean test13-pre4 w/ some netfilter fixes. ksymoops 0.7c on i686 2.4.0-test13-pre4. Options used -V (default) -K (specified) -L (specified) -o /lib/modules/2.4.0-test13-pre4/ (default) -m /usr/src/linux/System.map (default) No modules in ksyms, skipping objects Oops: CPU:0 EIP:0010:[] Using defaults from ksymoops -t elf32-i386 -a i386 EFLAGS: 00010082 eax: ebx: ecx: d5efc610 edx: d3efc610 esi: 0001 edi: c02fa0a4 ebp: 0003 esp: c0297e84 ds: 0018 es: 0018 ss: 0018 Process swapper (pid: 0, stackpage=c0297000) Stack: c02fa0a4 c019c088 c02fa0a4 d3efc610 d3efc650 0001 c02fa0a4 0003 c1578078 c019269e c019d2d5 c02fa0a4 0012 c019d260 c019d4d8 c02fa0a4 c02fa0a4 d3efc650 c0192a09 c02fa0a4 d3efc650 Call Trace: [<0019c08b>] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] [] Code: 83 78 0c 00 74 04 31 c0 eb 4a 8a 42 02 24 0f 0f b6 c0 83 f8 >>EIP; c019c017<= Trace; 0019c08b Before first symbol Trace; c019269e Trace; c019d2d5 Trace; c019d260 Trace; c019d4d8 Trace; c0192a09 Trace; c019c464 Trace; c019d173 Trace; c011ccd2 Trace; c0193127 Trace; c019d0ac Trace; c010a04f Trace; c010a1ac Trace; c01071f0 Trace; c01071f0 Trace; c0108e94 Trace; c01071f0 Trace; c01071f0 Trace; c0100018 Trace; c0107213 Trace; c0107279 Trace; c0105000 Trace; c0100191 Code; c019c017 <_EIP>: Code; c019c017<= 0: 83 78 0c 00 cmpl $0x0,0xc(%eax) <= Code; c019c01b 4: 74 04 je a <_EIP+0xa> c019c021 Code; c019c01d 6: 31 c0 xor%eax,%eax Code; c019c01f 8: eb 4a jmp54 <_EIP+0x54> c019c06b Code; c019c021 a: 8a 42 02 mov0x2(%edx),%al Code; c019c024 d: 24 0f and$0xf,%al Code; c019c026 f: 0f b6 c0 movzbl %al,%eax Code; c019c029 12: 83 f8 00 cmp$0x0,%eax "Mohammad A. Haque" wrote: > > I just captured the oops. > > It happens when you try to mount (mount -t udf /dev/foo /mnt/bar) an > encrypted dvd. At least it does on my end. Unencrypted dvds mount fine. > > ksymoops coming soon. > -- = Mohammad A. Haque http://www.haque.net/ [EMAIL PROTECTED] "Alcohol and calculus don't mix. Project Lead Don't drink and derive." --Unknown http://wm.themes.org/ [EMAIL PROTECTED] = - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: css hang; somewhere between test12 and test13pre4ac2
On Mon, Dec 25 2000, Mohammad A. Haque wrote: > This is confirmed. mounting css dics causes oops. non-css discs work > fine. > > oops coming soon. Thanks. If it's the cdrom_log_sense oops, please try also with previously sent patch. -- * Jens Axboe <[EMAIL PROTECTED]> * SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: css hang; somewhere between test12 and test13pre4ac2
This is confirmed. mounting css dics causes oops. non-css discs work fine. oops coming soon. Jens Axboe wrote: > The most likely suspect (as someone else pointed out) is not at > all css (I'm not even sure what you mean by css hang?) but UDF. > Given the fs changes. Since sysrq still works, it would help a > lot if you could capture sysrq-p repeatedly and send it in. > > Do you have any non-css discs to beat on UDF? > > -- > * Jens Axboe <[EMAIL PROTECTED]> > * SuSE Labs > - > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to [EMAIL PROTECTED] > Please read the FAQ at http://www.tux.org/lkml/ -- = Mohammad A. Haque http://www.haque.net/ [EMAIL PROTECTED] "Alcohol and calculus don't mix. Project Lead Don't drink and derive." --Unknown http://wm.themes.org/ [EMAIL PROTECTED] = - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: test13-pre4... udf problem with dvd access vs test12
I just captured the oops. It happens when you try to mount (mount -t udf /dev/foo /mnt/bar) an encrypted dvd. At least it does on my end. Unencrypted dvds mount fine. ksymoops coming soon. Frank Jacobberger wrote: > > Odd happening here. Been running good as gold through test12 with > accessing my dvd to using oms. Now updating to test13-pre4 > I get a complete lock up of my whole system when executing oms. > > I can access the drive via mounting it... with no trouble what ever. > > Here is a snip from my message file No clue what to test for here... > > Perhaps udf.c is the problem? > > Any ideas? -- = Mohammad A. Haque http://www.haque.net/ [EMAIL PROTECTED] "Alcohol and calculus don't mix. Project Lead Don't drink and derive." --Unknown http://wm.themes.org/ [EMAIL PROTECTED] = - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
BIOS problem, pro Microsoft, anti other OS
To Maintainer: PCI SUBSYSTEM P: Martin Mares M: [EMAIL PROTECTED] L: [EMAIL PROTECTED] S: Supported This alert should probably be forwarded to Others, but appropriate subTask persons in the kernel-source Maintainers list were not obvious. Briefly, documented below is the fact/complications that some PC BIOS chips are now coming with a default Microsoft setting, which makes them hostile to some functionalities of other OS. If particular under Linux, a PCI Winmodem did NOT function with the Win98 BIOS setting, but did fine with BIOS choice "Other OS". Possible, other PCI devices under Linux OS might be simmilarly afflicated. This indicates a need for Linux install software to be equipped with a utility to probe the BIOS and report back "Linux hostile" BIOS settings. Today most Newbies are getting new PC boxes equipped with WinModems. Hostile BIOS settings will block their capability to get on-line. Unfortunately, I do not have the technical capablity to directly contribute. Thus please forward this alert to however may be capable and concerned with dealing with the problem. MarvS, co-maintainer: http://walbran.org/sean/linux/linmodem-howto.html === Subject: Device or resource busy : SUCCESS ! Date: Sun, 24 Dec 2000 14:46:04 +0200 (IST) From: Jacques Goldberg <[EMAIL PROTECTED]> Reply-To: Jacques Goldberg <[EMAIL PROTECTED]> To: [EMAIL PROTECTED] Well, my very sincere thanks to all of you. It works. DETAILS: I purchased a Gateway Solo 2550 in September, comes with ActionTec PCI 56k modem (Lucent chip vendor 11c1, device 448). I tried the 568 ltmodem: device or resource busy,could not guess why. cat /pro/pci would show no interrupt Then ltmodem 578 was made available: same problem. But then the PnP issue at boot came again last week. I had tried several times to discover the option in my BIOS setup. This morning I found: in the "advanced" page there is an "Operating System" option, to be set to "the most frequently used OS". I had left it as Win98/2000 (as I received the machine). I just selected OTHER. Lo and behold, ltmodem.o loads without a flaw, I then had a short dumb terminal session with minicom, and am now connected at my first attempt with ppp, having already used X11, ssh, and Netscape. So, again my deepest thanks to all of you on this list, and the bottom line for newcomers: IF (Device.or.resource busy) CHECK YOUR BIOS. By the way I am running RH-6.1, kernel 2.2.12-20 (CERN "official" Linux distribution). They are on vacations now, I cannot check at this time if their version of ppp is "generic" or reworked. Jacques J. Goldberg [EMAIL PROTECTED] Currently at TECHNION PHONE: Technion=+(972)(0)(4)829.36.63 CERN=+(41)(22)767.84.72 Original Message Subject: Crippling BIOSes Date: Sun, 24 Dec 2000 14:55:36 -0500 From: Marvin Stodolsky <[EMAIL PROTECTED]> To: LinModems <[EMAIL PROTECTED]> Folks, Given Jacques report, it would be good to set up a diagnostic for BIOS which have such Microsoft/Other choices. Mine does not. For those of you who have such BIOSes, please 1) Do boots under both choices 2) Under microsoft do: dmesg > ms.txt 3) Under other dmesg > other.txt 4) diff ms.txt other.txt and report the differences to the List with the name of the BIOS. If would clearly be desirable to equip future Linux kernels/software to give a warning about the crippling microsoft option, which may hamper other PCI harware under Linux as well. MarvS Original Message Subject: Re: dmesg detection?? Date: Mon, 25 Dec 2000 09:29:20 +0200 (IST) From:Jacques Goldberg <[EMAIL PROTECTED]> Reply-To: Jacques Goldberg <[EMAIL PROTECTED]> To: Marvin Stodolsky <[EMAIL PROTECTED]> CC:[EMAIL PROTECTED] Gateway Solo 2150 Phoenix BIOS version 17.50 BIOS Page "Advanced" BIOS Field "Installed O/S" may be "Other" "Win98/Win2000" or "Win95" Did not try "Win95" Default was Win98/Win2000, dmesg file attached is ms.txt Changed to "Other" , dmesg file atached is other.txt Linux kernel 2.2.12-20 SOUND: CONFIG_SOUND set to "m" Using OSS driver (ES-1371 not supported by RH-6.1 2.2.12-20 distribution) Using PPP version 2.3.10-3 Everything works fine (ppp sessions with sound) once BIOS O/S choice set to OTHER. I repeat what my problem was: -could not load ltmodem : "Device or resource busy" -cat /pro/pci did not show IRQ, not even the word IRQ, for Lucent modem -setting BIOS to OTHER instantly made everything work. Jacques J. Goldberg [EMAIL PROTECTED] dmesg > ms.txt (Win98 BIOS setting) Linux version 2.2.12-20 ([EMAIL PROTECTED]) (gcc version
Re: Proposal: devfs names ending in %d or %u
On Sun, Dec 24, 2000 at 11:02:39PM -0800, Barry K. Nathan wrote: > Eric Shattow wrote: > [snip] > > when i insert a FAT formatted disc with a PC partition table, the partition > > i want to mount is part1. when i insert a HFS formatted disc with a MAC > > partition table, the partition i want to mount is part4. this is very ugly, > > and it has nothing to do with devfs. Those would be /dev/sda1 (adjust > device name for IDE instead of SCSI, etc.) and /dev/sda4 without devfs. > > In this case, the problem is that different Zip disks really do have their > data on different partitions. (If you use enough different disks and > formatting utilities, it won't even be the same partition for all PC disks > or all Mac disks, IIRC.) I don't use Zip disks much anymore, although > there's a similar phenomenon with my SCSI MO drive on my desktop Mac > (which I recently started using Linux on again). [snip] And don't forget that IDE zip drives can be programmed to hide the partition table altogether from the OS. Supposedly this can be changed via an ATAPI command to the drive, but I don't think the kernel driver has any support for this. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
[PATCH] CONFIG_MOUSE should not be tristate
Hello! CONFIG_MOUSE only enables further questions. It is never used except drivers/char/Config.in where it's checked for being "n". CONFIG_MOUSE=m makes no sence. The patch is against 2.4.0-test13-pre4. ___ --- linux.orig/drivers/char/Config.in +++ linux/drivers/char/Config.in @@ -95,7 +95,7 @@ fi fi -tristate 'Mouse Support (not serial and bus mice)' CONFIG_MOUSE +bool 'Mouse Support (not serial and bus mice)' CONFIG_MOUSE if [ "$CONFIG_MOUSE" != "n" ]; then bool ' PS/2 mouse (aka "auxiliary device") support' CONFIG_PSMOUSE tristate ' C 82C710 mouse port support (as on TI Travelmate)' CONFIG_82C710_MOUSE ___ Regards, Pavel Roskin - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Netgear FA311
Well now I've just set up the network and everything works a treat, currently getting 850kb/s file transfers. Cool. Except for one thing. In the logs I'm seeing this: Dec 25 15:25:18 tahallah last message repeated 2 times Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0783. Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0780. Dec 25 15:25:19 tahallah last message repeated 2 times Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0580. Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0783. Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0780. Dec 25 15:25:21 tahallah last message repeated 3 times Dec 25 15:25:21 tahallah kernel: eth0: Something Wicked happened! 0783. Dec 25 15:25:22 tahallah kernel: eth0: Something Wicked happened! 0780. That's just a small snippet. Apart from those messages, all seems OK, but I'm just wondering why it's getting those messages - what does those mean? Cheers, Alex -- Huffapuff! http://www.tahallah.clara.co.uk - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Masquerade hangups
On Sun, Dec 24, 2000 at 09:02:12AM -0600, Mike Elmore wrote: > Hello, > > > I seem to get pretty good performance from > internet->masq box and from masq box->internal > lan, but when a internal box tries to get to the > net through the masquerade, connection seem to time > out. I'll get a pretty good initial burst, then > connections stall. please join the netfilter/iptables mailinglist (instructions on http://netfilter.kernelnotes.org) and file us a detailed report. It's a good idea to save linux-kernel from all the nifty details :) > I'm using test13-pre4. I saw some iptables stuff on > the list a week or so ago, was this fixed in pre4 or > is this my problem? we (the netfilter core team) are currently not aware of any bugs at the moment. The behaviour you've described wasn't reported by anybody else. > -mwe > [EMAIL PROTECTED] -- Live long and prosper - Harald Welte / [EMAIL PROTECTED]http://www.gnumonks.org GCS/E/IT d- s-: a-- C+++ UL$ P+++ L$ E--- W- N++ o? K- w--- O- M- V-- PS+ PE-- Y+ PGP++ t++ 5-- !X !R tv-- b+++ DI? !D G+ e* h+ r% y+(*) - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Linux 2.4.0test13pre4ac2
On Mon, Dec 25, 2000 at 10:37:54PM +1100, Andrew Morton wrote: > Alan Cox wrote: > > > > 2.4.0test13pre4-ac2 > > > > o Make smp cpu halt synchronous (Andi Kleen) > > errr, Andi. > > We're asking all the other CPUs to call stop_this_cpu(), and then waiting > for them to complete the call. > > But stop_this_cpu() never returns, so the machine gets stuck. Right, it's wrong. Thanks for spotting that. -Andi - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: TCP keepalive seems to send to only one port
On Mon, Dec 25, 2000 at 04:27:07PM +0100, Igmar Palsenberg wrote: > > > Yeah. But I'm stuck with a NAT (which isn't mine, btw) which uses 2.1.xxx-2.2.x > > (according to nmap). Which had a default of 15 *minutes* (as I read in a HOWTO > > somewhere). I'm trying to convince the sysadmin to raise it to two hours, but I > > bet it'll be hard. > > ipchains -S timeoutval 0 0 is the only way to do this. > That's the easy part. Convincing the sysadmin (which probably is not very clueful) is the hard one. But this is getting too OT, EOT. -- Cesar Eduardo Barros [EMAIL PROTECTED] [EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: About Celeron processor memory barrier problem
[EMAIL PROTECTED] (Tim Wright) wrote on 24.12.00 in <[EMAIL PROTECTED]>: > On Sun, Dec 24, 2000 at 11:36:00AM +0200, Kai Henningsen wrote: > There was a similar thread to this recently. The issue is that if you > choose the wrong processor type, you may not even be able to complain. Hmm ... I think I can see ways around that (essentially similar to the 16 bit bootstrap code), but it may indeed be more trouble than it's worth. MfG Kai - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: TCP keepalive seems to send to only one port
> Yeah. But I'm stuck with a NAT (which isn't mine, btw) which uses 2.1.xxx-2.2.x > (according to nmap). Which had a default of 15 *minutes* (as I read in a HOWTO > somewhere). I'm trying to convince the sysadmin to raise it to two hours, but I > bet it'll be hard. ipchains -S timeoutval 0 0 is the only way to do this. Igmar - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
[PATCH] winbond-840 updates, tester needed!
I found a few minor problems in the current winbond driver: * the current driver hangs/crashes during module unload. * it doesn't use the new pci dma mapping interface * it assumed strong memory ordering without explicit wmb()'s. * it contains the tx_full race that Andrew Morton found. The patch fixes these bugs. I've tested it on i386 SMP. I need a tester with a winbond card on a computer with a big endian cpu, any volunteers? There are at least 2 outstanding problems with big endian cpus: * a warning from Donald Becker that one eeprom read is broken for big-endian machines. * the driver doesn't use the "Descriptor Big Endian" (bit 20 of register 0) mode, instead it uses cpu_to_le32. -- Manfred --- 2.4/drivers/net/winbond-840.c Sun Dec 17 18:03:56 2000 +++ build-2.4/drivers/net/winbond-840.c Mon Dec 25 12:49:36 2000 @@ -21,11 +21,21 @@ Do not change the version information unless an improvement has been made. Merely removing my name, as Compex has done in the past, does not count as an improvement. + + Changelog: + * ported to 2.4 + ??? + * spin lock update, memory barriers, new style dma mappings + Manfred Spraul + + TODO: + * according to the documentation, the chip supports big endian + internally. Replace the cpu_to_le32 with that bit. */ /* These identify the driver base version and may not be removed. */ static const char version1[] = -"winbond-840.c:v1.01 5/15/2000 Donald Becker <[EMAIL PROTECTED]>\n"; +"winbond-840.c:v1.01 (2.4 port) 5/15/2000 Donald Becker <[EMAIL PROTECTED]>\n"; static const char version2[] = " http://www.scyld.com/network/drivers.html\n"; @@ -114,12 +124,6 @@ #include #include -/* Condensed operations for readability. - The compatibility defines are in kern_compat.h */ - -#define virt_to_le32desc(addr) cpu_to_le32(virt_to_bus(addr)) -#define le32desc_to_virt(addr) bus_to_virt(le32_to_cpu(addr)) - MODULE_AUTHOR("Donald Becker <[EMAIL PROTECTED]>"); MODULE_DESCRIPTION("Winbond W89c840 Ethernet driver"); MODULE_PARM(max_interrupt_work, "i"); @@ -298,9 +302,12 @@ #define PRIV_ALIGN 15 /* Required alignment mask */ struct netdev_private { - /* Descriptor rings first for alignment. */ - struct w840_rx_desc rx_ring[RX_RING_SIZE]; - struct w840_tx_desc tx_ring[TX_RING_SIZE]; + struct w840_rx_desc *rx_ring; + dma_addr_t rx_addr[RX_RING_SIZE]; + struct w840_tx_desc *tx_ring; + dma_addr_t tx_addr[RX_RING_SIZE]; + dma_addr_t ring_dma_addr; + struct pci_dev *pdev; /* The addresses of receive-in-place skbuffs. */ struct sk_buff* rx_skbuff[RX_RING_SIZE]; /* The saved address of a sent-in-place packet/buffer, for later free(). */ @@ -335,7 +342,7 @@ static void check_duplex(struct net_device *dev); static void netdev_timer(unsigned long data); static void tx_timeout(struct net_device *dev); -static void init_ring(struct net_device *dev); +static int init_ring(struct net_device *dev); static int start_tx(struct sk_buff *skb, struct net_device *dev); static void intr_handler(int irq, void *dev_instance, struct pt_regs *regs); static void netdev_error(struct net_device *dev, int intr_status); @@ -364,6 +371,11 @@ return -EIO; pci_set_master(pdev); + if(!pci_dma_supported(pdev,0x)) { + printk(KERN_WARNING "Winbond-840: Device %s disabled due to DMA +limitations.\n", + pdev->name); + return -EIO; + } dev = init_etherdev(NULL, sizeof(*np)); if (!dev) return -ENOMEM; @@ -403,6 +415,7 @@ np = dev->priv; np->chip_id = chip_idx; np->drv_flags = pci_id_tbl[chip_idx].drv_flags; + np->pdev = pdev; spin_lock_init(>lock); pdev->driver_data = dev; @@ -632,10 +645,12 @@ printk(KERN_DEBUG "%s: w89c840_open() irq %d.\n", dev->name, dev->irq); - init_ring(dev); + if((i=init_ring(dev))) + return i; - writel(virt_to_bus(np->rx_ring), ioaddr + RxRingPtr); - writel(virt_to_bus(np->tx_ring), ioaddr + TxRingPtr); + writel(np->ring_dma_addr, ioaddr + RxRingPtr); + writel(np->ring_dma_addr+sizeof(struct w840_rx_desc)*RX_RING_SIZE, + ioaddr + TxRingPtr); for (i = 0; i < 6; i++) writeb(dev->dev_addr[i], ioaddr + StationAddr + i); @@ -733,11 +748,13 @@ "config %8.8x.\n", dev->name, (int)readl(ioaddr + IntrStatus), (int)readl(ioaddr + NetworkConfig)); + spin_lock_irq(>lock); check_duplex(dev); if (np->csr6 != old_csr6) { writel(np->csr6 & ~0x0002, ioaddr + NetworkConfig); writel(np->csr6 | 0x2002, ioaddr + NetworkConfig);
Re: css hang; somewhere between test12 and test13pre4ac2
On Mon, Dec 25 2000, Dave Gilbert wrote: > > The most likely suspect (as someone else pointed out) is not at > > all css (I'm not even sure what you mean by css hang?) but UDF. > > I mean a complete system hang when playing a CSS disc - doesn't even ping. > Doesn't recover. Hmm > > Given the fs changes. Since sysrq still works, it would help a > > lot if you could capture sysrq-p repeatedly and send it in. > > I think at this point the only thing that works is sysrq-b - at least the > sysrq-u's and sysrq-s's that I've given don't seem to have cleanly > unmounted the file system. Could you at least check? You may need some sort of serial console too.. > > Do you have any non-css discs to beat on UDF? > > Yep one disc (Scanners) - it is fine - hence my reason for beleiving it is > a CSSism (although I guess CSS makes other demands on the UDF code). Not so. Once a css "session" has been established, data is read just like off any other CD. But try with this patch applied, it could be a NULL pointer deref at the wrong time. -- * Jens Axboe <[EMAIL PROTECTED]> * SuSE Labs --- drivers/ide/ide-cd.c~ Sat Dec 23 23:59:52 2000 +++ drivers/ide/ide-cd.cSun Dec 24 00:03:38 2000 @@ -333,7 +333,7 @@ { int log = 0; - if (sense == NULL || pc->quiet) + if (sense == NULL || pc == NULL || pc->quiet) return 0; switch (sense->sense_key) {
Re: css hang; somewhere between test12 and test13pre4ac2
On Mon, 25 Dec 2000, Jens Axboe wrote: > The most likely suspect (as someone else pointed out) is not at > all css (I'm not even sure what you mean by css hang?) but UDF. I mean a complete system hang when playing a CSS disc - doesn't even ping. Doesn't recover. > Given the fs changes. Since sysrq still works, it would help a > lot if you could capture sysrq-p repeatedly and send it in. I think at this point the only thing that works is sysrq-b - at least the sysrq-u's and sysrq-s's that I've given don't seem to have cleanly unmounted the file system. > Do you have any non-css discs to beat on UDF? Yep one disc (Scanners) - it is fine - hence my reason for beleiving it is a CSSism (although I guess CSS makes other demands on the UDF code). Dave -- Have a happy GNU millennium! -- / Dr. David Alan Gilbert | Running GNU/Linux on | Happy \ \ gro.gilbert @ treblig.org | Alpha, x86, ARM and SPARC | In Hex / \ ___|___ http://www.treblig.org |/ - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
2.2.18 compile warnings
I get a large number of warnings like the following when compiling 2.2.18: {standard input}:338: Warning: using `%eax' instead of `%ax' due to `l' suffix I'm using the compiler from RedHat 6.2: Reading specs from /usr/lib/gcc-lib/i386-redhat-linux/egcs-2.91.66/specs gcc version egcs-2.91.66 19990314/Linux (egcs-1.1.2 release) - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
new linux_logo16
Since the 16-color logo was broken a while ago, we need a new one for 2.4.0. The main limitation is that we no longer can choose the palette, but have to use the standard VGA 16-color palette. I quantized the 256-color logo to the VGA palette using ppmquant and used my rudimentary artistical skills with the GIMP to remove the background pattern. The results are attached (yes, they are small): - new_logo16.png: plain quantized 256-color logo - new_logo16_fs.png: quantized 256-color logo using Floyd-Steinberg dithering - new_logo16_2.png: new_logo16.png with the background pattern removed So either some artistic soul creates a new non-offensive politically-correct[*] logo using the standard VGA 16-color palette, or I'll submit new_logo16_2.png for inclusion in 2.4.0. Any comments? Gr{oetje,eeting}s, Geert [*] I don't want another to-drink-or-not-to-drink discussion. -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [EMAIL PROTECTED] In personal conversations with technical people, I call myself a hacker. But when I'm talking to journalists I just say "programmer" or something like that. -- Linus Torvalds new_logo16.png new_logo16_fs.png new_logo16_2.png
Re: Linux 2.4.0test13pre4ac2
Alan Cox wrote: > > 2.4.0test13pre4-ac2 > > o Make smp cpu halt synchronous (Andi Kleen) errr, Andi. We're asking all the other CPUs to call stop_this_cpu(), and then waiting for them to complete the call. But stop_this_cpu() never returns, so the machine gets stuck. What were you trying to do here, BTW? - - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: css hang; somewhere between test12 and test13pre4ac2
On Sun, Dec 24 2000, Dave Gilbert wrote: > Hi, > Somewhere between test12 and test13pre4ac2 (sheesh the version > numbers.) CSS on ATAPI DVD ROM drives has stopped working. > > Playing a CSS disc (using xine) causes a complete system hang (machine > doesn't ping - sysrq-b still works) on test13pre4ac2. On test12 it is > still OK. > > This is on an Alpha LX164. The most likely suspect (as someone else pointed out) is not at all css (I'm not even sure what you mean by css hang?) but UDF. Given the fs changes. Since sysrq still works, it would help a lot if you could capture sysrq-p repeatedly and send it in. Do you have any non-css discs to beat on UDF? -- * Jens Axboe <[EMAIL PROTECTED]> * SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Driver for emulating a tape device on top of a cd writer...
On Tue, Dec 19 2000, Stelian Pop wrote: > > > Basically, I would like to be able to use a cdwriter as a tape > > > device, with software like dump(8) or tar(1). With /dev/tcdw > > > as name (for example), I'd like to be able to do: > > > [...] > > > What you describe is actually one of the goals of the packet writing > > driver. To do this reliably you need packet writing, I won't even > > start to think about the headaches wihtout it... > > Yes, I saw your patch for packet writing but: > - the CD written with packet writing software may not be readable > on standard CD-ROM drives (and I want that, because almost > everybody has one). On CD drives sold during the last two years or so, and of course all DVD drives they are readable. But of course of you want 100% coverage, it isn't good enough. > - using packet writing you basically write _files_ on top of an > UDF filesystem. Tar and dump (or afio, cpio etc) does not > support that kind of access, they expect to be given a character > device they can stream data to. (Of course, it is possible to > add some additionnal level of indirection on top of the packet > device and provide character based access to the UDF files, but > IMHO _this_ would be overkill). Why would you even want to use UDF for this? You want raw access to the device. Packet writing or not, this is totally unrelated. > - data backups are expected to be fast. Writing data in DAO/TAO > mode is much quicker than in packet mode. No no no, not much quicker. Write large packets and it's just as fast as dao/tao. 64Kb packets are a bit slower because of run-in, run-out block over head, but using larger packets this isn't the noticable. And packet writing has so many other advantages... > - reliability is a question of implementation. cdrecord can > be very reliable. If a user space application can provide this > level of reliability, it should be even simpler to achieve it > in kernel space (and I plan to use the BurnProof/etc extensions > which will be present on all future cdwriters). Even simpler to achieve reliability in the kernel? I gather you mean feeding-data reliability, and not stability. > > > I'll start to work on this, probably by looking at the cdrecord > > > low level code and porting it into kernel space. > > > > Oh god no! You can do all this from user space. > > Please pay attention to the fact that I was refering to the 'low level > code'. I don't intend to write a driver who can replace cdrecord. > _This_ would be madness. Very much so > What I indend to do is just a 'small' driver, which supports only the > mmc drives. I expect the driver to be only some hundreds lines long. A few hundred lines? *This* I look forward to seeing :) > Doing that from user space would mean propagating the data from > the user space application (dump or tar) to a character mode > driver, and back to a user space application (something like a hacked > cdrecord), which will return in kernel space using sg interface... > It could be easier to write (even if I don't exactly feel confident > about hacking the cdrecord source :) ), but the reliability and > the performance would be far far away... Pipes and 100% user space based, then pass to sg? I don't see the problem. -- * Jens Axboe <[EMAIL PROTECTED]> * SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
test13-pre4... udf problem with dvd access vs test12
Odd happening here. Been running good as gold through test12 with accessing my dvd to using oms. Now updating to test13-pre4 I get a complete lock up of my whole system when executing oms. I can access the drive via mounting it... with no trouble what ever. Here is a snip from my message file No clue what to test for here... Perhaps udf.c is the problem? Any ideas? Merry Christmas Frank Dec 25 02:02:13 f1j kernel: hdf: packet command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:13 f1j kernel: hdf: packet command error: error=0x00 Dec 25 02:02:13 f1j kernel: ATAPI device hdf: Dec 25 02:02:13 f1j kernel: Error: No sense data -- (Sense key=0x00) Dec 25 02:02:13 f1j kernel: No additional sense information -- (asc=0x00, ascq=0x00) Dec 25 02:02:13 f1j kernel: The failed "Report Key" packet command was: Dec 25 02:02:13 f1j kernel: "a4 00 00 00 00 00 00 00 00 0c c4 00 " Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: end_request: I/O error, dev 21:40 (hdf), sector 1148 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: end_request: I/O error, dev 21:40 (hdf), sector 1152 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: end_request: I/O error, dev 21:40 (hdf), sector 1153 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: end_request: I/O error, dev 21:40 (hdf), sector 1154 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: end_request: I/O error, dev 21:40 (hdf), sector 1155 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: innd mmap bug in 2.4.0-test12
On Mon, 25 Dec 2000, Linus Torvalds wrote: > > Assuming we don't lose any PG_dirty bits, we might of course just lose it > from the page tables themselves before it ever even gets to "struct page". > I'm just surprised that it seems to be so repeatable for you - it sounds > like we _never_ actually write out the dirty pages to disk. It's not that > we can lose the dirty bit occasionally, we seem to lose it every time in > your setup. Nope. I got it. The thing is even more embarrassing than just losing a dirty bit. We don't lose any dirty bits (well, we could before, but after adding the PageDirty() test to reclaim_page() we're ok now). In fact, we know _exactly_ which pages are dirty, and which pages are not. We just don't write them out. Because right now the only thing that writes out dirty pages is memory pressure. "sync()", "fsync()" and "fdatasync()" will happily ignore dirty pages completely. The thing that made me overlook that simple thing in testing was that I was testing the new VM stuff under heavy VM load - to shake out any bugs. Under heavy VM load, there are no problems, because the memory pressure will make sure everything gets written out. Under heavy VM load the thing works just beautifully. Under _low_, or no, memory pressure, however, the dang thing just stays in memory. We'll happily reboot with the new contents still cached, in fact. I bet that if you start something that eats up all your memory, and causes some nice swapping just before you shut down the machine, your innd active file will be right as rain after a reboot. I'm a stupid git. I even remember thinking about the syncing issues at some point, and then obviously just forgetting about it _completely_. The simple fix is along the lines of adding code to fsync() that walks the inode page list and writes out dirty pages. The clever and clean fix is to split the inode page list into two lists, one for dirty and one for clean pages, and only walk the dirty list. Ho ho ho. I _so_ enjoy making a fool out of myself. Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: innd mmap bug in 2.4.0-test12
On Mon, 25 Dec 2000, Marco d'Itri wrote: > On Dec 24, Linus Torvalds <[EMAIL PROTECTED]> wrote: > > >/* The page is dirty, or locked, move to inactive_dirty list. */ > >if (page->buffers || TryLockPage(page)) { > >... > > > >and change the test to > > > >if (page->buffers || PageDirty(page) || TryLockPage(page)) { > Done, no change. > Got some articles, restarted the server, all is good. > Got other articles, rebooted and the files now differ. Willing to test some more? Add a printk() to __remove_inode_page() that complains whenever it removes a dirty page. Oh, in order to not see this with swap pages (which _can_ be removed when they are dirty, if all users of them are gone), add a PageClearDirty() to "remove_from_swap_cache()" so that we don't get false positives.. Do you get any messages? I don't think you will, but it should be tested. You might mark it a BUG(), so tht we'll get a stack-trace if it happens. Assuming we don't lose any PG_dirty bits, we might of course just lose it from the page tables themselves before it ever even gets to "struct page". I'm just surprised that it seems to be so repeatable for you - it sounds like we _never_ actually write out the dirty pages to disk. It's not that we can lose the dirty bit occasionally, we seem to lose it every time in your setup. I wonder if it's something specific innd does. Like "msync()" just being broken or similar. But the code looks sane. Hmm.. Can you send me an "strace" of innd when this happens? > And I have another problem: I'm experiencing random hangs using X[1] with > 2.4.0-test12. That's probably the infinite loop in the tty task queue handling, should be fixed in test13-pre3 or so. Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: innd mmap bug in 2.4.0-test12
On Mon, 25 Dec 2000, Marco d'Itri wrote: On Dec 24, Linus Torvalds [EMAIL PROTECTED] wrote: /* The page is dirty, or locked, move to inactive_dirty list. */ if (page-buffers || TryLockPage(page)) { ... and change the test to if (page-buffers || PageDirty(page) || TryLockPage(page)) { Done, no change. Got some articles, restarted the server, all is good. Got other articles, rebooted and the files now differ. Willing to test some more? Add a printk() to __remove_inode_page() that complains whenever it removes a dirty page. Oh, in order to not see this with swap pages (which _can_ be removed when they are dirty, if all users of them are gone), add a PageClearDirty() to "remove_from_swap_cache()" so that we don't get false positives.. Do you get any messages? I don't think you will, but it should be tested. You might mark it a BUG(), so tht we'll get a stack-trace if it happens. Assuming we don't lose any PG_dirty bits, we might of course just lose it from the page tables themselves before it ever even gets to "struct page". I'm just surprised that it seems to be so repeatable for you - it sounds like we _never_ actually write out the dirty pages to disk. It's not that we can lose the dirty bit occasionally, we seem to lose it every time in your setup. I wonder if it's something specific innd does. Like "msync()" just being broken or similar. But the code looks sane. Hmm.. Can you send me an "strace" of innd when this happens? And I have another problem: I'm experiencing random hangs using X[1] with 2.4.0-test12. That's probably the infinite loop in the tty task queue handling, should be fixed in test13-pre3 or so. Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: innd mmap bug in 2.4.0-test12
On Mon, 25 Dec 2000, Linus Torvalds wrote: Assuming we don't lose any PG_dirty bits, we might of course just lose it from the page tables themselves before it ever even gets to "struct page". I'm just surprised that it seems to be so repeatable for you - it sounds like we _never_ actually write out the dirty pages to disk. It's not that we can lose the dirty bit occasionally, we seem to lose it every time in your setup. Nope. I got it. The thing is even more embarrassing than just losing a dirty bit. We don't lose any dirty bits (well, we could before, but after adding the PageDirty() test to reclaim_page() we're ok now). In fact, we know _exactly_ which pages are dirty, and which pages are not. We just don't write them out. Because right now the only thing that writes out dirty pages is memory pressure. "sync()", "fsync()" and "fdatasync()" will happily ignore dirty pages completely. The thing that made me overlook that simple thing in testing was that I was testing the new VM stuff under heavy VM load - to shake out any bugs. Under heavy VM load, there are no problems, because the memory pressure will make sure everything gets written out. Under heavy VM load the thing works just beautifully. Under _low_, or no, memory pressure, however, the dang thing just stays in memory. We'll happily reboot with the new contents still cached, in fact. I bet that if you start something that eats up all your memory, and causes some nice swapping just before you shut down the machine, your innd active file will be right as rain after a reboot. I'm a stupid git. I even remember thinking about the syncing issues at some point, and then obviously just forgetting about it _completely_. The simple fix is along the lines of adding code to fsync() that walks the inode page list and writes out dirty pages. The clever and clean fix is to split the inode page list into two lists, one for dirty and one for clean pages, and only walk the dirty list. Ho ho ho. I _so_ enjoy making a fool out of myself. Linus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
test13-pre4... udf problem with dvd access vs test12
Odd happening here. Been running good as gold through test12 with accessing my dvd to using oms. Now updating to test13-pre4 I get a complete lock up of my whole system when executing oms. I can access the drive via mounting it... with no trouble what ever. Here is a snip from my message file No clue what to test for here... Perhaps udf.c is the problem? Any ideas? Merry Christmas Frank Dec 25 02:02:13 f1j kernel: hdf: packet command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:13 f1j kernel: hdf: packet command error: error=0x00 Dec 25 02:02:13 f1j kernel: ATAPI device hdf: Dec 25 02:02:13 f1j kernel: Error: No sense data -- (Sense key=0x00) Dec 25 02:02:13 f1j kernel: No additional sense information -- (asc=0x00, ascq=0x00) Dec 25 02:02:13 f1j kernel: The failed "Report Key" packet command was: Dec 25 02:02:13 f1j kernel: "a4 00 00 00 00 00 00 00 00 0c c4 00 " Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: end_request: I/O error, dev 21:40 (hdf), sector 1148 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: end_request: I/O error, dev 21:40 (hdf), sector 1152 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: end_request: I/O error, dev 21:40 (hdf), sector 1153 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: end_request: I/O error, dev 21:40 (hdf), sector 1154 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) Dec 25 02:02:14 f1j kernel: hdf: command error: status=0x51 { DriveReady SeekComplete Error } Dec 25 02:02:14 f1j kernel: hdf: command error: error=0x50 Dec 25 02:02:14 f1j kernel: end_request: I/O error, dev 21:40 (hdf), sector 1155 Dec 25 02:02:14 f1j kernel: ATAPI device hdf: Dec 25 02:02:14 f1j kernel: Error: Illegal request -- (Sense key=0x05) Dec 25 02:02:14 f1j kernel: Read of scrambled sector without authentication -- (asc=0x6f, ascq=0x03) - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: css hang; somewhere between test12 and test13pre4ac2
On Sun, Dec 24 2000, Dave Gilbert wrote: Hi, Somewhere between test12 and test13pre4ac2 (sheesh the version numbers.) CSS on ATAPI DVD ROM drives has stopped working. Playing a CSS disc (using xine) causes a complete system hang (machine doesn't ping - sysrq-b still works) on test13pre4ac2. On test12 it is still OK. This is on an Alpha LX164. The most likely suspect (as someone else pointed out) is not at all css (I'm not even sure what you mean by css hang?) but UDF. Given the fs changes. Since sysrq still works, it would help a lot if you could capture sysrq-p repeatedly and send it in. Do you have any non-css discs to beat on UDF? -- * Jens Axboe [EMAIL PROTECTED] * SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Driver for emulating a tape device on top of a cd writer...
On Tue, Dec 19 2000, Stelian Pop wrote: Basically, I would like to be able to use a cdwriter as a tape device, with software like dump(8) or tar(1). With /dev/tcdw as name (for example), I'd like to be able to do: [...] What you describe is actually one of the goals of the packet writing driver. To do this reliably you need packet writing, I won't even start to think about the headaches wihtout it... Yes, I saw your patch for packet writing but: - the CD written with packet writing software may not be readable on standard CD-ROM drives (and I want that, because almost everybody has one). On CD drives sold during the last two years or so, and of course all DVD drives they are readable. But of course of you want 100% coverage, it isn't good enough. - using packet writing you basically write _files_ on top of an UDF filesystem. Tar and dump (or afio, cpio etc) does not support that kind of access, they expect to be given a character device they can stream data to. (Of course, it is possible to add some additionnal level of indirection on top of the packet device and provide character based access to the UDF files, but IMHO _this_ would be overkill). Why would you even want to use UDF for this? You want raw access to the device. Packet writing or not, this is totally unrelated. - data backups are expected to be fast. Writing data in DAO/TAO mode is much quicker than in packet mode. No no no, not much quicker. Write large packets and it's just as fast as dao/tao. 64Kb packets are a bit slower because of run-in, run-out block over head, but using larger packets this isn't the noticable. And packet writing has so many other advantages... - reliability is a question of implementation. cdrecord can be very reliable. If a user space application can provide this level of reliability, it should be even simpler to achieve it in kernel space (and I plan to use the BurnProof/etc extensions which will be present on all future cdwriters). Even simpler to achieve reliability in the kernel? I gather you mean feeding-data reliability, and not stability. I'll start to work on this, probably by looking at the cdrecord low level code and porting it into kernel space. Oh god no! You can do all this from user space. Please pay attention to the fact that I was refering to the 'low level code'. I don't intend to write a driver who can replace cdrecord. _This_ would be madness. Very much so What I indend to do is just a 'small' driver, which supports only the mmc drives. I expect the driver to be only some hundreds lines long. A few hundred lines? *This* I look forward to seeing :) Doing that from user space would mean propagating the data from the user space application (dump or tar) to a character mode driver, and back to a user space application (something like a hacked cdrecord), which will return in kernel space using sg interface... It could be easier to write (even if I don't exactly feel confident about hacking the cdrecord source :) ), but the reliability and the performance would be far far away... Pipes and 100% user space based, then pass to sg? I don't see the problem. -- * Jens Axboe [EMAIL PROTECTED] * SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
2.2.18 compile warnings
I get a large number of warnings like the following when compiling 2.2.18: {standard input}:338: Warning: using `%eax' instead of `%ax' due to `l' suffix I'm using the compiler from RedHat 6.2: Reading specs from /usr/lib/gcc-lib/i386-redhat-linux/egcs-2.91.66/specs gcc version egcs-2.91.66 19990314/Linux (egcs-1.1.2 release) - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: css hang; somewhere between test12 and test13pre4ac2
On Mon, 25 Dec 2000, Jens Axboe wrote: The most likely suspect (as someone else pointed out) is not at all css (I'm not even sure what you mean by css hang?) but UDF. I mean a complete system hang when playing a CSS disc - doesn't even ping. Doesn't recover. Given the fs changes. Since sysrq still works, it would help a lot if you could capture sysrq-p repeatedly and send it in. I think at this point the only thing that works is sysrq-b - at least the sysrq-u's and sysrq-s's that I've given don't seem to have cleanly unmounted the file system. Do you have any non-css discs to beat on UDF? Yep one disc (Scanners) - it is fine - hence my reason for beleiving it is a CSSism (although I guess CSS makes other demands on the UDF code). Dave -- Have a happy GNU millennium! -- / Dr. David Alan Gilbert | Running GNU/Linux on | Happy \ \ gro.gilbert @ treblig.org | Alpha, x86, ARM and SPARC | In Hex / \ ___|___ http://www.treblig.org |/ - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: css hang; somewhere between test12 and test13pre4ac2
On Mon, Dec 25 2000, Dave Gilbert wrote: The most likely suspect (as someone else pointed out) is not at all css (I'm not even sure what you mean by css hang?) but UDF. I mean a complete system hang when playing a CSS disc - doesn't even ping. Doesn't recover. Hmm Given the fs changes. Since sysrq still works, it would help a lot if you could capture sysrq-p repeatedly and send it in. I think at this point the only thing that works is sysrq-b - at least the sysrq-u's and sysrq-s's that I've given don't seem to have cleanly unmounted the file system. Could you at least check? You may need some sort of serial console too.. Do you have any non-css discs to beat on UDF? Yep one disc (Scanners) - it is fine - hence my reason for beleiving it is a CSSism (although I guess CSS makes other demands on the UDF code). Not so. Once a css "session" has been established, data is read just like off any other CD. But try with this patch applied, it could be a NULL pointer deref at the wrong time. -- * Jens Axboe [EMAIL PROTECTED] * SuSE Labs --- drivers/ide/ide-cd.c~ Sat Dec 23 23:59:52 2000 +++ drivers/ide/ide-cd.cSun Dec 24 00:03:38 2000 @@ -333,7 +333,7 @@ { int log = 0; - if (sense == NULL || pc-quiet) + if (sense == NULL || pc == NULL || pc-quiet) return 0; switch (sense-sense_key) {
[PATCH] winbond-840 updates, tester needed!
I found a few minor problems in the current winbond driver: * the current driver hangs/crashes during module unload. * it doesn't use the new pci dma mapping interface * it assumed strong memory ordering without explicit wmb()'s. * it contains the tx_full race that Andrew Morton found. The patch fixes these bugs. I've tested it on i386 SMP. I need a tester with a winbond card on a computer with a big endian cpu, any volunteers? There are at least 2 outstanding problems with big endian cpus: * a warning from Donald Becker that one eeprom read is broken for big-endian machines. * the driver doesn't use the "Descriptor Big Endian" (bit 20 of register 0) mode, instead it uses cpu_to_le32. -- Manfred --- 2.4/drivers/net/winbond-840.c Sun Dec 17 18:03:56 2000 +++ build-2.4/drivers/net/winbond-840.c Mon Dec 25 12:49:36 2000 @@ -21,11 +21,21 @@ Do not change the version information unless an improvement has been made. Merely removing my name, as Compex has done in the past, does not count as an improvement. + + Changelog: + * ported to 2.4 + ??? + * spin lock update, memory barriers, new style dma mappings + Manfred Spraul + + TODO: + * according to the documentation, the chip supports big endian + internally. Replace the cpu_to_le32 with that bit. */ /* These identify the driver base version and may not be removed. */ static const char version1[] = -"winbond-840.c:v1.01 5/15/2000 Donald Becker [EMAIL PROTECTED]\n"; +"winbond-840.c:v1.01 (2.4 port) 5/15/2000 Donald Becker [EMAIL PROTECTED]\n"; static const char version2[] = " http://www.scyld.com/network/drivers.html\n"; @@ -114,12 +124,6 @@ #include asm/bitops.h #include asm/io.h -/* Condensed operations for readability. - The compatibility defines are in kern_compat.h */ - -#define virt_to_le32desc(addr) cpu_to_le32(virt_to_bus(addr)) -#define le32desc_to_virt(addr) bus_to_virt(le32_to_cpu(addr)) - MODULE_AUTHOR("Donald Becker [EMAIL PROTECTED]"); MODULE_DESCRIPTION("Winbond W89c840 Ethernet driver"); MODULE_PARM(max_interrupt_work, "i"); @@ -298,9 +302,12 @@ #define PRIV_ALIGN 15 /* Required alignment mask */ struct netdev_private { - /* Descriptor rings first for alignment. */ - struct w840_rx_desc rx_ring[RX_RING_SIZE]; - struct w840_tx_desc tx_ring[TX_RING_SIZE]; + struct w840_rx_desc *rx_ring; + dma_addr_t rx_addr[RX_RING_SIZE]; + struct w840_tx_desc *tx_ring; + dma_addr_t tx_addr[RX_RING_SIZE]; + dma_addr_t ring_dma_addr; + struct pci_dev *pdev; /* The addresses of receive-in-place skbuffs. */ struct sk_buff* rx_skbuff[RX_RING_SIZE]; /* The saved address of a sent-in-place packet/buffer, for later free(). */ @@ -335,7 +342,7 @@ static void check_duplex(struct net_device *dev); static void netdev_timer(unsigned long data); static void tx_timeout(struct net_device *dev); -static void init_ring(struct net_device *dev); +static int init_ring(struct net_device *dev); static int start_tx(struct sk_buff *skb, struct net_device *dev); static void intr_handler(int irq, void *dev_instance, struct pt_regs *regs); static void netdev_error(struct net_device *dev, int intr_status); @@ -364,6 +371,11 @@ return -EIO; pci_set_master(pdev); + if(!pci_dma_supported(pdev,0x)) { + printk(KERN_WARNING "Winbond-840: Device %s disabled due to DMA +limitations.\n", + pdev-name); + return -EIO; + } dev = init_etherdev(NULL, sizeof(*np)); if (!dev) return -ENOMEM; @@ -403,6 +415,7 @@ np = dev-priv; np-chip_id = chip_idx; np-drv_flags = pci_id_tbl[chip_idx].drv_flags; + np-pdev = pdev; spin_lock_init(np-lock); pdev-driver_data = dev; @@ -632,10 +645,12 @@ printk(KERN_DEBUG "%s: w89c840_open() irq %d.\n", dev-name, dev-irq); - init_ring(dev); + if((i=init_ring(dev))) + return i; - writel(virt_to_bus(np-rx_ring), ioaddr + RxRingPtr); - writel(virt_to_bus(np-tx_ring), ioaddr + TxRingPtr); + writel(np-ring_dma_addr, ioaddr + RxRingPtr); + writel(np-ring_dma_addr+sizeof(struct w840_rx_desc)*RX_RING_SIZE, + ioaddr + TxRingPtr); for (i = 0; i 6; i++) writeb(dev-dev_addr[i], ioaddr + StationAddr + i); @@ -733,11 +748,13 @@ "config %8.8x.\n", dev-name, (int)readl(ioaddr + IntrStatus), (int)readl(ioaddr + NetworkConfig)); + spin_lock_irq(np-lock); check_duplex(dev); if (np-csr6 != old_csr6) { writel(np-csr6 ~0x0002, ioaddr + NetworkConfig); writel(np-csr6 | 0x2002, ioaddr + NetworkConfig);
Re: TCP keepalive seems to send to only one port
Yeah. But I'm stuck with a NAT (which isn't mine, btw) which uses 2.1.xxx-2.2.x (according to nmap). Which had a default of 15 *minutes* (as I read in a HOWTO somewhere). I'm trying to convince the sysadmin to raise it to two hours, but I bet it'll be hard. ipchains -S timeoutval 0 0 is the only way to do this. Igmar - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: About Celeron processor memory barrier problem
[EMAIL PROTECTED] (Tim Wright) wrote on 24.12.00 in [EMAIL PROTECTED]: On Sun, Dec 24, 2000 at 11:36:00AM +0200, Kai Henningsen wrote: There was a similar thread to this recently. The issue is that if you choose the wrong processor type, you may not even be able to complain. Hmm ... I think I can see ways around that (essentially similar to the 16 bit bootstrap code), but it may indeed be more trouble than it's worth. MfG Kai - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: TCP keepalive seems to send to only one port
On Mon, Dec 25, 2000 at 04:27:07PM +0100, Igmar Palsenberg wrote: Yeah. But I'm stuck with a NAT (which isn't mine, btw) which uses 2.1.xxx-2.2.x (according to nmap). Which had a default of 15 *minutes* (as I read in a HOWTO somewhere). I'm trying to convince the sysadmin to raise it to two hours, but I bet it'll be hard. ipchains -S timeoutval 0 0 is the only way to do this. That's the easy part. Convincing the sysadmin (which probably is not very clueful) is the hard one. But this is getting too OT, EOT. -- Cesar Eduardo Barros [EMAIL PROTECTED] [EMAIL PROTECTED] - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Linux 2.4.0test13pre4ac2
On Mon, Dec 25, 2000 at 10:37:54PM +1100, Andrew Morton wrote: Alan Cox wrote: 2.4.0test13pre4-ac2 o Make smp cpu halt synchronous (Andi Kleen) errr, Andi. We're asking all the other CPUs to call stop_this_cpu(), and then waiting for them to complete the call. But stop_this_cpu() never returns, so the machine gets stuck. Right, it's wrong. Thanks for spotting that. -Andi - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Masquerade hangups
On Sun, Dec 24, 2000 at 09:02:12AM -0600, Mike Elmore wrote: Hello, I seem to get pretty good performance from internet-masq box and from masq box-internal lan, but when a internal box tries to get to the net through the masquerade, connection seem to time out. I'll get a pretty good initial burst, then connections stall. please join the netfilter/iptables mailinglist (instructions on http://netfilter.kernelnotes.org) and file us a detailed report. It's a good idea to save linux-kernel from all the nifty details :) I'm using test13-pre4. I saw some iptables stuff on the list a week or so ago, was this fixed in pre4 or is this my problem? we (the netfilter core team) are currently not aware of any bugs at the moment. The behaviour you've described wasn't reported by anybody else. -mwe [EMAIL PROTECTED] -- Live long and prosper - Harald Welte / [EMAIL PROTECTED]http://www.gnumonks.org GCS/E/IT d- s-: a-- C+++ UL$ P+++ L$ E--- W- N++ o? K- w--- O- M- V-- PS+ PE-- Y+ PGP++ t++ 5-- !X !R tv-- b+++ DI? !D G+ e* h+ r% y+(*) - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Netgear FA311
Well now I've just set up the network and everything works a treat, currently getting 850kb/s file transfers. Cool. Except for one thing. In the logs I'm seeing this: Dec 25 15:25:18 tahallah last message repeated 2 times Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0783. Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0780. Dec 25 15:25:19 tahallah last message repeated 2 times Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0580. Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0783. Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0780. Dec 25 15:25:21 tahallah last message repeated 3 times Dec 25 15:25:21 tahallah kernel: eth0: Something Wicked happened! 0783. Dec 25 15:25:22 tahallah kernel: eth0: Something Wicked happened! 0780. That's just a small snippet. Apart from those messages, all seems OK, but I'm just wondering why it's getting those messages - what does those mean? Cheers, Alex -- Huffapuff! http://www.tahallah.clara.co.uk - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
[PATCH] CONFIG_MOUSE should not be tristate
Hello! CONFIG_MOUSE only enables further questions. It is never used except drivers/char/Config.in where it's checked for being "n". CONFIG_MOUSE=m makes no sence. The patch is against 2.4.0-test13-pre4. ___ --- linux.orig/drivers/char/Config.in +++ linux/drivers/char/Config.in @@ -95,7 +95,7 @@ fi fi -tristate 'Mouse Support (not serial and bus mice)' CONFIG_MOUSE +bool 'Mouse Support (not serial and bus mice)' CONFIG_MOUSE if [ "$CONFIG_MOUSE" != "n" ]; then bool ' PS/2 mouse (aka "auxiliary device") support' CONFIG_PSMOUSE tristate ' CT 82C710 mouse port support (as on TI Travelmate)' CONFIG_82C710_MOUSE ___ Regards, Pavel Roskin - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Proposal: devfs names ending in %d or %u
On Sun, Dec 24, 2000 at 11:02:39PM -0800, Barry K. Nathan wrote: Eric Shattow wrote: [snip] when i insert a FAT formatted disc with a PC partition table, the partition i want to mount is part1. when i insert a HFS formatted disc with a MAC partition table, the partition i want to mount is part4. this is very ugly, and it has nothing to do with devfs. Those would be /dev/sda1 (adjust device name for IDE instead of SCSI, etc.) and /dev/sda4 without devfs. In this case, the problem is that different Zip disks really do have their data on different partitions. (If you use enough different disks and formatting utilities, it won't even be the same partition for all PC disks or all Mac disks, IIRC.) I don't use Zip disks much anymore, although there's a similar phenomenon with my SCSI MO drive on my desktop Mac (which I recently started using Linux on again). [snip] And don't forget that IDE zip drives can be programmed to hide the partition table altogether from the OS. Supposedly this can be changed via an ATAPI command to the drive, but I don't think the kernel driver has any support for this. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
BIOS problem, pro Microsoft, anti other OS
To Maintainer: PCI SUBSYSTEM P: Martin Mares M: [EMAIL PROTECTED] L: [EMAIL PROTECTED] S: Supported This alert should probably be forwarded to Others, but appropriate subTask persons in the kernel-source Maintainers list were not obvious. Briefly, documented below is the fact/complications that some PC BIOS chips are now coming with a default Microsoft setting, which makes them hostile to some functionalities of other OS. If particular under Linux, a PCI Winmodem did NOT function with the Win98 BIOS setting, but did fine with BIOS choice "Other OS". Possible, other PCI devices under Linux OS might be simmilarly afflicated. This indicates a need for Linux install software to be equipped with a utility to probe the BIOS and report back "Linux hostile" BIOS settings. Today most Newbies are getting new PC boxes equipped with WinModems. Hostile BIOS settings will block their capability to get on-line. Unfortunately, I do not have the technical capablity to directly contribute. Thus please forward this alert to however may be capable and concerned with dealing with the problem. MarvS, co-maintainer: http://walbran.org/sean/linux/linmodem-howto.html === Subject: Device or resource busy : SUCCESS ! Date: Sun, 24 Dec 2000 14:46:04 +0200 (IST) From: Jacques Goldberg [EMAIL PROTECTED] Reply-To: Jacques Goldberg [EMAIL PROTECTED] To: [EMAIL PROTECTED] Well, my very sincere thanks to all of you. It works. DETAILS: I purchased a Gateway Solo 2550 in September, comes with ActionTec PCI 56k modem (Lucent chip vendor 11c1, device 448). I tried the 568 ltmodem: device or resource busy,could not guess why. cat /pro/pci would show no interrupt Then ltmodem 578 was made available: same problem. But then the PnP issue at boot came again last week. I had tried several times to discover the option in my BIOS setup. This morning I found: in the "advanced" page there is an "Operating System" option, to be set to "the most frequently used OS". I had left it as Win98/2000 (as I received the machine). I just selected OTHER. Lo and behold, ltmodem.o loads without a flaw, I then had a short dumb terminal session with minicom, and am now connected at my first attempt with ppp, having already used X11, ssh, and Netscape. So, again my deepest thanks to all of you on this list, and the bottom line for newcomers: IF (Device.or.resource busy) CHECK YOUR BIOS. By the way I am running RH-6.1, kernel 2.2.12-20 (CERN "official" Linux distribution). They are on vacations now, I cannot check at this time if their version of ppp is "generic" or reworked. Jacques J. Goldberg [EMAIL PROTECTED] Currently at TECHNION PHONE: Technion=+(972)(0)(4)829.36.63 CERN=+(41)(22)767.84.72 Original Message Subject: Crippling BIOSes Date: Sun, 24 Dec 2000 14:55:36 -0500 From: Marvin Stodolsky [EMAIL PROTECTED] To: LinModems [EMAIL PROTECTED] Folks, Given Jacques report, it would be good to set up a diagnostic for BIOS which have such Microsoft/Other choices. Mine does not. For those of you who have such BIOSes, please 1) Do boots under both choices 2) Under microsoft do: dmesg ms.txt 3) Under other dmesg other.txt 4) diff ms.txt other.txt and report the differences to the List with the name of the BIOS. If would clearly be desirable to equip future Linux kernels/software to give a warning about the crippling microsoft option, which may hamper other PCI harware under Linux as well. MarvS Original Message Subject: Re: dmesg detection?? Date: Mon, 25 Dec 2000 09:29:20 +0200 (IST) From:Jacques Goldberg [EMAIL PROTECTED] Reply-To: Jacques Goldberg [EMAIL PROTECTED] To: Marvin Stodolsky [EMAIL PROTECTED] CC:[EMAIL PROTECTED] Gateway Solo 2150 Phoenix BIOS version 17.50 BIOS Page "Advanced" BIOS Field "Installed O/S" may be "Other" "Win98/Win2000" or "Win95" Did not try "Win95" Default was Win98/Win2000, dmesg file attached is ms.txt Changed to "Other" , dmesg file atached is other.txt Linux kernel 2.2.12-20 SOUND: CONFIG_SOUND set to "m" Using OSS driver (ES-1371 not supported by RH-6.1 2.2.12-20 distribution) Using PPP version 2.3.10-3 Everything works fine (ppp sessions with sound) once BIOS O/S choice set to OTHER. I repeat what my problem was: -could not load ltmodem : "Device or resource busy" -cat /pro/pci did not show IRQ, not even the word IRQ, for Lucent modem -setting BIOS to OTHER instantly made everything work. Jacques J. Goldberg [EMAIL PROTECTED] dmesg ms.txt (Win98 BIOS setting) Linux version 2.2.12-20 ([EMAIL PROTECTED]) (gcc version egcs-2.91.66 19990314/Linux
Re: css hang; somewhere between test12 and test13pre4ac2
On Mon, Dec 25 2000, Mohammad A. Haque wrote: This is confirmed. mounting css dics causes oops. non-css discs work fine. oops coming soon. Thanks. If it's the cdrom_log_sense oops, please try also with previously sent patch. -- * Jens Axboe [EMAIL PROTECTED] * SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: css hang; somewhere between test12 and test13pre4ac2
Heh, just go it. Testing Jens Axboe wrote: Thanks. If it's the cdrom_log_sense oops, please try also with previously sent patch. -- = Mohammad A. Haque http://www.haque.net/ [EMAIL PROTECTED] "Alcohol and calculus don't mix. Project Lead Don't drink and derive." --Unknown http://wm.themes.org/ [EMAIL PROTECTED] = - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Fatal Oops on boot with 2.4.0testX and recent GCC snapshots
Hello Mike, hello linux-kernel hackers, Mike Galbraith wrote: I wouldn't (not going to here;) spend a lot of time on it. The compiler has problems. It won't build glibc-2.2, and chokes horribly on ipchains. Maybe, but you were lucky getting an ICE, and not silently failing code :-) After having spent several hours debugging now, I think it was worth it (at least for my understanding of lower-level kernel issues and of the (rather nice and almost readable) assembly code gcc generates). There seems to be something going wrong in the down(sem) path after the kernel_thread call. I'm not sure if down() succeeds instantly when compiling the kernel with 2.95.2, but it seems to fail for 2.97; I figured out by spilling some printk's around in bdflush_init, which made the bug magically disappear, due to the looser timing. This also might happen for compiling with frame pointers or with the static declaration variables, somehow. Th bdflush_init function itself does not seem to be responsible, which corresponds with the assembly, which is fine and should get the same results for all compiled cases. It seems that whyever, the cause for this failure is actually the down(sem) call on a not yet up()'ed semaphore, and this is where it starts to get ugly. down() then calls __down_failed, which ends up in __down(); __down does some waitqueue handling, which I don't understand, and then calls __wake_up - up to then, everything seems fine, in __wake_up it is where my search ended up to now, but I think something is wrong in this context; however, the complexity of this code exceeds my knowledge by magnitudes, so I can't continue searching there without going mad :-) It would be nice if someone else could look from there on, now I've narrowed the case down to rather low-level functions. Greetings, Andreas -- Andreas Franck - --- [EMAIL PROTECTED] -- - Keep smiling! - - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Fatal Oops on boot with 2.4.0testX and recent GCC snapshots
Hello Mike, hello linux-kernel hackers, Mike Galbraith wrote: I wouldn't (not going to here;) spend a lot of time on it. The compiler has problems. It won't build glibc-2.2, and chokes horribly on ipchains. Maybe, but after having spent several hours debugging now, I think it was worth it: I am almost sure this is not a gcc bug, but a nasty race condition involving the semaphore handling bdflush_init. I figured out by spilling some printk's around in bdflush_init, which made the bug magically disappear, what wasn't what I intended - but which gave me a clearer impression of what's going on. It seems that whyever, the cause for this failure is actually the down(sem) call on a not yet up()'ed semaphore, and this is where it starts to get ugly. -- Andreas Franck - --- [EMAIL PROTECTED] -- - Keep smiling! - - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: BIOS problem, pro Microsoft, anti other OS
Hello! This alert should probably be forwarded to Others, but appropriate subTask persons in the kernel-source Maintainers list were not obvious. Briefly, documented below is the fact/complications that some PC BIOS chips are now coming with a default Microsoft setting, which makes them hostile to some functionalities of other OS. If particular under Linux, a PCI Winmodem did NOT function with the Win98 BIOS setting, but did fine with BIOS choice "Other OS". Possible, other PCI devices under Linux OS might be simmilarly afflicated. This indicates a need for Linux install software to be equipped with a utility to probe the BIOS and report back "Linux hostile" BIOS settings. Today most Newbies are getting new PC boxes equipped with WinModems. Hostile BIOS settings will block their capability to get on-line. Unfortunately, I do not have the technical capablity to directly contribute. Thus please forward this alert to however may be capable and concerned with dealing with the problem. Can you check what does Linux 2.4.0-testlatest behave, please? I know of these problems and I hope the new PCI code in 2.4.0 is able to assign the missing memory/IO resources without help of the BIOS, but unfortunately 2.2 isn't and it's very difficult to back-port the fixes as they depend on changes in many other parts of the kernel. You probably should make the ltmodem driver check the region base registers and interrupts and if they are not set, recommend the user to change the OS or PNP settings in their BIOS setup. Have a nice fortnight -- Martin `MJ' Mares [EMAIL PROTECTED] [EMAIL PROTECTED] http://atrey.karlin.mff.cuni.cz/~mj/ First law of socio-genetics: Celibacy is not hereditary. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: test13-pre4... udf problem with dvd access vs test12
On Mon, Dec 25 2000, Mohammad A. Haque wrote: Jens, I made sure to reverse the udf patch I mentioned in another thread (all it really is merging changes from linux-udf cvs into the current kernel). So this is from a clean test13-pre4 w/ some netfilter fixes. Ok, looks unrelated however. EIP; c019c017 cdrom_log_sense+f/68 = Yes I know about this one, I've attached the patch here again. Linus, could you apply? --- drivers/ide/ide-cd.c~ Sat Dec 23 23:59:52 2000 +++ drivers/ide/ide-cd.cSun Dec 24 00:03:38 2000 @@ -333,7 +333,7 @@ { int log = 0; - if (sense == NULL || pc-quiet) + if (sense == NULL || pc == NULL || pc-quiet) return 0; switch (sense-sense_key) { -- * Jens Axboe [EMAIL PROTECTED] * SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: test13-pre4... udf problem with dvd access vs test12
Fix confirmed. Am i supposed to get some DriveSense errors? I probably am just don't recall. Jens Axboe wrote: Yes I know about this one, I've attached the patch here again. Linus, could you apply? --- drivers/ide/ide-cd.c~ Sat Dec 23 23:59:52 2000 +++ drivers/ide/ide-cd.cSun Dec 24 00:03:38 2000 @@ -333,7 +333,7 @@ { int log = 0; - if (sense == NULL || pc-quiet) + if (sense == NULL || pc == NULL || pc-quiet) return 0; switch (sense-sense_key) { -- * Jens Axboe [EMAIL PROTECTED] * SuSE Labs -- = Mohammad A. Haque http://www.haque.net/ [EMAIL PROTECTED] "Alcohol and calculus don't mix. Project Lead Don't drink and derive." --Unknown http://wm.themes.org/ [EMAIL PROTECTED] = - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: test13-pre4... udf problem with dvd access vs test12
Sorry, I meant DriveReady SeekComplete Error "Mohammad A. Haque" wrote: Fix confirmed. Am i supposed to get some DriveSense errors? I probably am just don't recall. -- = Mohammad A. Haque http://www.haque.net/ [EMAIL PROTECTED] "Alcohol and calculus don't mix. Project Lead Don't drink and derive." --Unknown http://wm.themes.org/ [EMAIL PROTECTED] = - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
serial mouse - lockup connection 2.4.0-t12
greetings! the lockup of test12 leaves no droppings i can find, but in the course of a half-dozen lockups in the last few days i've made an observation or two that may be of diagnostic help. system is a k6-2 on a via chipset mb with onboard everything but video (fic va-503a). kernel built with gcc-2.95.2. glibc is 2.2. the lockups occur during mouse movement -- mouse is kensington emulating microsoft. the peculiar thing is that the mouse acceleration seems to slow a little in the seconds before the lockup. i don't know that this is a useful datum, but it seemed as if it might be significant. the lockups are otherwise neither predictable or reproducible. -- dep -- bipartisanship: an illogical construct not unlike the idea that if half the people like red and half the people like blue, the country's favorite color is purple. - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: test13-pre4... udf problem with dvd access vs test12
On Mon, Dec 25 2000, Mohammad A. Haque wrote: Fix confirmed. Am i supposed to get some DriveSense errors? I probably am just don't recall. Good. The whole idea of cdrom_log_sense is to be able to cleanly limit the (often) sense-less (yes, haha :) ide-cd verbosity. Some of the css stuff are obvious candidates. We want users to be aware of the error, but often we end up flooding the logs with the same stuff over and over again. This is often a source of confusion for the casual user. -- * Jens Axboe [EMAIL PROTECTED] * SuSE Labs - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Netgear FA311
Alex wrote: In the logs I'm seeing this: Dec 25 15:25:18 tahallah last message repeated 2 times Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0783. Dec 25 15:25:19 tahallah kernel: eth0: Something Wicked happened! 0780. 783 means: Tx Underrun Tx Idle Tx Packet Error Tx Descriptor Rx Packet Error Rx Descriptor Rx OK. Hmm. I download the Documentation from National (http://www.national.com/pf/DP/DP83815.html), and the the tx burst size/fill threshold/drain threshold combination is invalid: from natsemi.c: /* Configure the PCI bus bursts and FIFO thresholds. */ /* Configure for standard, in-spec Ethernet. */ np-tx_config = (128) + /* Automatic transmit padding */ (123) + /* Excessive collision retry */ (0x020) + /* Max DMA burst = 512 byte */ (88) +/* fill threshold = 256 byte */ 2; /* drain threshold = 64 byte */ writel(np-tx_config, ioaddr + TxConfig); But: page 51 The MXDMA MUST NOT be greater than the Tx Fill Threshold Could you try this setup? /* Configure the PCI bus bursts and FIFO thresholds. */ /* Configure for standard, in-spec Ethernet. */ np-tx_config = (128) + /* Automatic transmit padding */ (123) + /* Excessive collision retry */ (620) + /* Max DMA burst = 128 byte */ (88) +/* fill threshold = 256 byte */ 8; /* drain threshold = 256 byte */ writel(np-tx_config, ioaddr + TxConfig); -- Manfred - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Netgear FA311
On Mon, 25 Dec 2000, Manfred wrote: Could you try this setup? /* Configure the PCI bus bursts and FIFO thresholds. */ /* Configure for standard, in-spec Ethernet. */ np-tx_config = (128) + /* Automatic transmit padding */ (123) + /* Excessive collision retry */ (620) + /* Max DMA burst = 128 byte */ (88) +/* fill threshold = 256 byte */ 8; /* drain threshold = 256 byte */ writel(np-tx_config, ioaddr + TxConfig); Hmm, that little change worked a lot better. However thoughput is down to 700kb/s! Transferring files from the other machine to this machine is much faster - 868kb/s. In the logs, I only got *one* message from the natsemi driver (and this happened when sending files from this machine to the other machine. No problems receiving from the other machine. Dec 25 17:28:12 tahallah kernel: eth0: Something Wicked happened! 0583. But I just realised that the other machine I'm using has an 10 megabit ethernet card (on the hub that one is shown on 10, my machine is shown as 100). I think this explains the throughput problem. When I get another 4 way power socket, I'll put my other machine (which has a 100 megabit card in it) on the network and see if that makes a difference. Cheers, Alex -- Huffapuff! http://www.tahallah.clara.co.uk - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Fatal Oops on boot with 2.4.0testX and recent GCC snapshots
On Mon, 25 Dec 2000, Andreas Franck wrote: Hello Mike, hello linux-kernel hackers, Mike Galbraith wrote: I wouldn't (not going to here;) spend a lot of time on it. The compiler has problems. It won't build glibc-2.2, and chokes horribly on ipchains. Maybe, but after having spent several hours debugging now, I think it was worth it: I am almost sure this is not a gcc bug, but a nasty race condition involving the semaphore handling bdflush_init. I figured out by spilling some printk's around in bdflush_init, which made the bug magically disappear, what wasn't what I intended - but which gave me a clearer impression of what's going on. Oh? Can you show me (offline) what you did exactly that made it go away? (that's kinda scary.. _much_ prefer 'compiler has rough edges' option;) -Mike - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: css hang; somewhere between test12 and test13pre4ac2
Hi Jens, Thanks for the patch; unfortunatly at the moment I can neither confirm or deny whether the patch works - since the damn thing won't crash on me with or without the patch - I suspect it depends on system mood. But I have a PDA logging console so if it happens I should catch it. Thanks again, Dave -- Have a happy GNU millennium! -- / Dr. David Alan Gilbert | Running GNU/Linux on | Happy \ \ gro.gilbert @ treblig.org | Alpha, x86, ARM and SPARC | In Hex / \ ___|___ http://www.treblig.org |/ - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: Fatal Oops on boot with 2.4.0testX and recent GCC snapshots
On Mon, 25 Dec 2000, Andreas Franck wrote: Hello Mike, hello linux-kernel hackers, Mike Galbraith wrote: I wouldn't (not going to here;) spend a lot of time on it. The compiler has problems. It won't build glibc-2.2, and chokes horribly on ipchains. Maybe, but you were lucky getting an ICE, and not silently failing code :-) You bet. After having spent several hours debugging now, I think it was worth it (at least for my understanding of lower-level kernel issues and of the (rather nice and almost readable) assembly code gcc generates). There Don't get me wrong, chasing things like this is never a waste of time. In the case of gcc in particular. Our next 'stable' kernel compiler is going to come from the gcc development tree just as the next 'stable' kernel is coming out of the kernel development tree. -Mike - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: BIOS problem, pro Microsoft, anti other OS
Jacques Can you check what does Linux 2.4.0-testlatest behave, please? Are you comfortable compiling kernels? If not I'll compile one for you for the test. Currently I'm happily under 2.4.0-test 12: lsmod: Module Size Used by ppp_deflate39164 1 (autoclean) bsd_comp4148 0 (autoclean) ppp_async 6220 1 (autoclean) ppp_generic12820 2 (autoclean) [ppp_deflate bsd_comp ppp_async] ltmodem 364948 1 input 3328 0 serial 42192 1 (autoclean) (ONLY suporting my mouse) isa-pnp27528 0 (autoclean) [ltmodem serial] usbcore27684 0 (unused) Martin Are there any particular kernel config choices that will be beneficial for this problem? Mark You probably should make the ltmodem driver check the region base registers and interrupts and if they are not set, recommend the user to change the OS or PNP settings in their BIOS setup. Can these be included in the LTmodem packages you are working up? MarvS - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/
Re: innd mmap bug in 2.4.0-test12
On Dec 25, Linus Torvalds [EMAIL PROTECTED] wrote: Add a printk() to __remove_inode_page() that complains whenever it removes a dirty page. Oh, in order to not see this with swap pages (which _can_ be removed when they are dirty, if all users of them are gone), add a PageClearDirty() to "remove_from_swap_cache()" so that we don't get false positives.. Do you get any messages? I don't think you will, but it should be tested. I read you found the real cause so that may be bogus, but I have got two messages while booting. The first showed up while doing the fsck of a 6 GB file systems and killed the process (fscks of smaller partitions completed successfully), the second occured while initializing /dev/random and left an unkillable dd process and a stuck boot process (I gathered this info with sysrq). Being -test12 unstable for me, if you don't need more data I'll go back to -test9 until the next release. That's probably the infinite loop in the tty task queue handling, should be fixed in test13-pre3 or so. Looks like I missed it, evil vger postmasters unsubscribed me again for no apparent reason... -- ciao, Marco - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] Please read the FAQ at http://www.tux.org/lkml/