Re: Sticky Keys ?
You want accessx for X-windows. Solaris, Compaq/Digital, and SGI provide it, but I didn't see anything at www.xfree86.org Searching around the web found a version for Linux http://slappy.cs.uiuc.edu/fall98/Linux/download.html AccessX appears to have been developed, at least in part, by the Trace Research and Development Center at the University of Wisconsin-Madison. The page at http://trace.wisc.edu/docs/x_win_access/x_access.htm says, in the audience question section: Audience: Will AccessX be in X11R6? Will: AccessX was developed by Digital, Trace, and SUN and we did it for X11R5. Digital and SUN will be releasing it with the next version of their operating systems which include the X11R5 server. DACX has worked closely with Silicon Graphics, who are working to develop the XKB extension for X11R6. It is a much larger extension that deals with the keyboard and is a logical place to put the AccessX code. So for X11R6, AccessX will be a part of a larger extension called XKB. The answer is yes it will be there but it will be a different name. Earl: One additional item. The fact that the underlying is changing, AccessX to XKB, the user interface will still be the same, so to use the R5 version on a SUN or DEC and when you transition to R6 version of the window server, the user interface will look the same, the interaction will be the same. So you won't have to change how you interact with the systems. XKB is part of X11R6.1 and later, so I infer that the low-level X server support for sticky keys is built into recent versions of XFree86. I don't know whether the UI stuff to control it is part of X11R6.x, or of XFree86, though. The Linux AccessX page you cite has a tarball of "the pristine source" for their package, which appears to contain a Tcl script which may be their control application, plus some C and C++ source; that page seems to imply that there are features over and above the XKB-based features of the Digital/Trace/Sun AccessX project, e.g.: o Video Mode Changing lets users change their video screen mode on demand. o Control Panel allows the user to apply the settings before saving, save the user's settings, tab through the panel (for those who cannot use a mouse), give the user the option to restore the to the default settings, and more. o Soon, the AccessX package will also include screen magnification. (I don't know whether the Video Mode Changing lets you change resolution on the fly without changing your desktop - in which case it'd probably be of interest even to people *without* limited vision - or not; it may just be an interface to the video-mode changing extensions of XFree86). To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: invlpg produces strange sig11 on PentiumPro box
On 01-Jul-00 Poul-Henning Kamp wrote: In message [EMAIL PROTECTED], Mike Smith writes: In message [EMAIL PROTECTED], KATO Takenori writ es: The invlpg instruction causes strange signal 11 problem on some PentiumPro box. This problem seems to hapen when (1) mother board is very old and (2) BIOS update is not available and (3) cpuid 0x619. Please Mike, just because you see my name you shouldn't take a contrary positition until you have actually looked into matters. Look at the first paragraph: This is for Pentium Pro cpus running in motherboards where the BIOS does not contain the needed microcode updates. *ahem* You might want to read the first paragraph as well. It is for situations where one _can't_ update one's BIOS. I don't see why making it a tweakable kernel compile time know that is off by default would be so incredibly bad. We have precedents already for this type of thing. And yes, in this case, the CPU is not performing as advertised. -- John Baldwin [EMAIL PROTECTED] -- http://www.FreeBSD.org/~jhb/ PGP Key: http://www.cslab.vt.edu/~jobaldwi/pgpkey.asc "Power Users Use the Power to Serve!" - http://www.FreeBSD.org/ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: UDF (DVD fs)
It seems Coleman Kane wrote: Hello, is anyone currently working on code to implement the UDF filesystem? For those not familiar with it, it is the filesystem that DVDs use. I'd like to look into getting the support under FreeBSD, since the players already seem to work. If no one is working on this, then I could probably use some help in writing the code to support this fs. I think Julian Elischer is working on UDF.. However to play/read/use DVD's you dont need UDF, they are also readable as an ISO9660, but that might change in the future... -Søren To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: UDF (DVD fs)
I am working on UDF support. I have at present a program that reads a udf filesystem and am working (today) on making it into an "mtools" like program that allows access to the contents in a useful manner. I will eventually turn this into a (readonly) filesystem, and it is designed with that in mind (it uses a buffer cache etc, like the kernel. (in other words I'm prototyping). I will at some stage also try make a UDF creation module for mkisofs as well. In the meanwhile you should be able to mount most modern DVDs using the ISO9660 filesystem as they should be "bridge" format, (in which there is metadata for both types of filesystems). Julian Coleman Kane wrote: Hello, is anyone currently working on code to implement the UDF filesystem? For those not familiar with it, it is the filesystem that DVDs use. I'd like to look into getting the support under FreeBSD, since the players already seem to work. If no one is working on this, then I could probably use some help in writing the code to support this fs. -- Coleman Kane President, UC Free O.S. Users Group - http://pohl.ececs.uc.edu --- Part 1.2 Type: application/pgp-signature -- __--_|\ Julian Elischer / \ [EMAIL PROTECTED] ( OZ) World tour 2000 )_.---._/ presently in: Budapest v To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Why do we always create a malloc disk for md?
In message [EMAIL PROTECTED], John Baldwin writes: This ends up registering md1 with disk_create, and thus md1 is returned as a disk through kern.disks into the list returned by Disk_Names(), and sysinstall blows up when it tries to open it. I think the reason it blows up is because /dev/md1 isn't around, although I think I may be able to fix that by adding 'md' as a disk device in the table in sysinstall/devices.c.\ However, I'm curious if md1 should be created in this case? Your diagnosis is correct. sysinstall should ignore md* disks. The reason md1 is created is that it is the beginning of an "almost-clone" implementation. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 [EMAIL PROTECTED] | TCP/IP since RFC 956 FreeBSD coreteam member | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: invlpg produces strange sig11 on PentiumPro box
In message [EMAIL PROTECTED], John Baldwin writes: *ahem* You might want to read the first paragraph as well. It is for situations where one _can't_ update one's BIOS. I don't see why making it a tweakable kernel compile time know that is off by default would be so incredibly bad. We have precedents already for this type of thing. And yes, in this case, the CPU is not performing as advertised. So far we have set the limit at hardware being used correctly. Either way, this patch was not the correct way to fix this particular erratum. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 [EMAIL PROTECTED] | TCP/IP since RFC 956 FreeBSD coreteam member | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: UDF (DVD fs)
It seems Julian Elischer wrote: I am working on UDF support. I have at present a program that reads a udf filesystem and am working (today) on making it into an "mtools" like program that allows access to the contents in a useful manner. I will eventually turn this into a (readonly) filesystem, and it is designed with that in mind (it uses a buffer cache etc, like the kernel. (in other words I'm prototyping). I will at some stage also try make a UDF creation module for mkisofs as well. Uhm, the real value of UDF is that it can be used as a "real" rw filesystem on CDRW/DVDRAM media, if this is not implemented the value of having UDF is very limited IMHO In the meanwhile you should be able to mount most modern DVDs using the ISO9660 filesystem as they should be "bridge" format, (in which there is metadata for both types of filesystems). Endeed, makeing a ro UDF filesystem more or less useless :) -Søren To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: vmstats for pages that go inactive-active
On Sat, 1 Jul 2000, I wrote: vmmeter-cnt.v_reactivated counts the number of cache pages that get promoted to either active or inactive queues. My read (and I could be wrong) from vm/vm_page.c is, there is no statistic to count the inactive pages that get "reclaimed" into the active queue. Another idea would be to simply include the number of reclaimed inactive pages into the original cnt.v_reactivated stats, since they are in this sense basically the same as reclaimed cache pages. Thoughts? -Paul. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
RE: Invalidating PACK!!!
what do you mean by ribbon cables, aren't the cables that came with the motherboard good for it? -Original Message- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On Behalf Of Joey Miller Sent: Sat, July 01, 2000 10:04 PM To: Mustafa Deeb Cc: [EMAIL PROTECTED]; [EMAIL PROTECTED] Subject: Re: Invalidating PACK!!! had the same problems. make sure you are using LVD ribbon cables. On Sat, Jul 01, 2000 at 11:11:47AM +0200, Mustafa Deeb wrote: // hi all, // // we build our own servers, we've always used the intel N440BX and the // Barracuda disks.. and we liked it so much, // this time we bought intel's L440GX+ and the chettah disks from seagate // (Ultra2 DIsks) // and I'm getting these errors, // ofcourse the server goes nuts when an error like this happens.. // after looking into the mailling lists, nobody gave a direct reason for this // problem or even a solution, anyways, I want to replace the Hard Drives, // // is there someone who've used the L440GX+ motherboard with 18G disks and he // is happy with it // // Best Regards... // Mustafa N. Deeb // // // // da0 at ahc0 bus 0 target 1 lun 0 // da0: SEAGATE ST318404LW 0002 Fixed Direct Access SCSI-3 device // da0: 20.000MB/s transfers (10.000MHz, offset 63, 16bit), Tagged Queueing // Enabled // da0: 17501MB (35843670 512 byte sectors: 255H 63S/T 2231C) // da2 at ahc0 bus 0 target 4 lun 0 // da2: SEAGATE ST318404LW 0002 Fixed Direct Access SCSI-3 device // da2: 20.000MB/s transfers (10.000MHz, offset 63, 16bit), Tagged Queueing // Enabled // da2: 17501MB (35843670 512 byte sectors: 255H 63S/T 2231C) // da1 at ahc0 bus 0 target 2 lun 0 // da1: SEAGATE ST318404LW 0002 Fixed Direct Access SCSI-3 device // da1: 20.000MB/s transfers (10.000MHz, offset 63, 16bit), Tagged Queueing // Enabled // da1: 17501MB (35843670 512 byte sectors: 255H 63S/T 2231C) // // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): SCB 0x62 - timed out while idle, SEQADDR == 0xb // (da2:ahc0:0:4:0): Queuing a BDR SCB // (da2:ahc0:0:4:0): no longer in timeout, status = 34a // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): Invalidating pack // (da2:ahc0:0:4:0): SCB 0xc - timed out while idle, SEQADDR == 0x9 // (da2:ahc0:0:4:0): Queuing a BDR SCB // (da2:ahc0:0:4:0): no longer in timeout, status = 34a // (da2:ahc0:0:4:0): Invalidating pack // // // // To Unsubscribe: send mail to [EMAIL PROTECTED] // with "unsubscribe freebsd-isp" in the body of the message -- Joey Miller Sr. Systems Engineer iBIZ Technology Corp. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: UDF (DVD fs)
Soren Schmidt wrote: It seems Julian Elischer wrote: Indeed, makeing a ro UDF filesystem more or less useless :) I will be adding read/write later.. One must be able to walk before running! As for RO UDF, it is useful for reading UDF2.0.1 cds that don't have ISO9660 on them (I have one here). The experience of looking at how the other vendors do read/write (especially adaptec) in order to be able to read their disks will be used to make it writable. -Søren -- __--_|\ Julian Elischer / \ [EMAIL PROTECTED] ( OZ) World tour 2000 )_.---._/ presently in: Budapest v To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Documentation selection in sysinstall
2. We need to find a UI model that allows the user to efficiently select the language and formats they want to install. I'm thinking of initially presenting a dialog box that looks like this: Documentation is available in the following languages: English Spanish French Japanese Chinese with the list extending as necessary, based on what sysinstall found on the FTP site. After the user has chosen a language, then present them with a list like this: Now choose the documentation you would like to install, and the formats you would like to use. HTML HTMLText PS PDF PDB RTF Split Books Handbook[ ][X] [ ][ ] [ ] [ ] [ ] FAQ [X][X] [ ][ ] [ ] [ ] [ ] Porter's Handbook [ ][ ] [ ][X] [ ] [ ] [ ] [... and so on ...] You actually have a sparsly populated three dimensional array here. Documentation[ material, language, format ] Exmample: Documentation[ Handbook, EN, HTML ] Documentation[ Handbook, EN, HTMLSplit ] Documentation[ Handbook, JA, HTMLSplit ] ... Suggestion: Use a LIST of preferences for language: ja,de,...,en use the first one that you find in order of preference, default is 'en' ... ??? You specify the language preference Preferred language: en[*] ja[1] de[2] (Don't ask me how you do this with the current sysinstall tools!) Suggestion: Select the set of media/format you want HTML Split HTMLText PS PDF PDB RTF [x][x][ ] [ ] [ ] [ ] [ ] This is pretty simple to set up and do, and is similar to how a lot of other documentation packages are structured. Only my opinion of course. Patrick To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: style(9)
Warner Losh wrote: In message [EMAIL PROTECTED] Jeroen Ruigrok/Asmodai writes: : Often the 80 column boundary reminds me not to use : functions_which_have_crazy_long_names_with_underscores(), but be a : little more brief, but not too. ;) IKnowPeopleThatLikeToHaveParagraphFunctionNamesToo(); DrivesMeNutsBecauseTheirCodeIsHardToReadAndModify(); TooMuchVerbosityIsSoMuchWorseThanTooLittleAndRepeitionCanBeBadToo(); stop(THE_INSANITY); -- "Where am I, and what am I doing in this handbasket?" Wes Peters Softweyr LLC [EMAIL PROTECTED] http://softweyr.com/ To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: style(9)
In message [EMAIL PROTECTED] Wes Peters writes: : Warner Losh wrote: : : In message [EMAIL PROTECTED] Jeroen Ruigrok/Asmodai :writes: : : Often the 80 column boundary reminds me not to use : : functions_which_have_crazy_long_names_with_underscores(), but be a : : little more brief, but not too. ;) : : IKnowPeopleThatLikeToHaveParagraphFunctionNamesToo(); : DrivesMeNutsBecauseTheirCodeIsHardToReadAndModify(); : TooMuchVerbosityIsSoMuchWorseThanTooLittleAndRepeitionCanBeBadToo(); : : stop(THE_INSANITY); SIGILL. Terminated. Warner To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
name cache size
Hi, for the purposes of an experiment, how can I increase the size of freebsd's name cache? Currently, find -type f on a tree with more than 40,000 files refuses to cache the results, and increasing desiredvnodes and the size of nchashtbl didn't seem to help. (4.0-RELEASE, with maxusers=512). thanks in advance, (should this have gone to -questions instead? wasn't sure..) -- Sitaram Iyer [EMAIL PROTECTED]http://www.cs.rice.edu/~ssiyer/ Phone: (off) 713-348-2265 (rice xBANK), (home) 713-799-8980 To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
dev to name
Hi. How can i find out the name of device file by device major/minor? To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: dev to name
On Sunday, July 02, 2000, Fox Anderson wrote: Hi. How can i find out the name of device file by device major/minor? The devtoname() function. ``man devtoname'' -- |Chris Costello [EMAIL PROTECTED] |To define recursion, we must first define recursion. ` To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: name cache size
* Sitaram Iyer [EMAIL PROTECTED] [000702 12:32] wrote: Hi, for the purposes of an experiment, how can I increase the size of freebsd's name cache? Currently, find -type f on a tree with more than 40,000 files refuses to cache the results, and increasing desiredvnodes and the size of nchashtbl didn't seem to help. (4.0-RELEASE, with maxusers=512). thanks in advance, (should this have gone to -questions instead? wasn't sure..) Try this: sysctl -w vfs.vmiodirenable=1 -Alfred To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: name cache size
Thus, Alfred Perlstein wrote... how can I increase the size of freebsd's name cache? sysctl -w vfs.vmiodirenable=1 thanks, that worked. -- Sitaram Iyer [EMAIL PROTECTED]http://www.cs.rice.edu/~ssiyer/ Phone: (off) 713-348-2265 (rice xBANK), (home) 713-799-8980 To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
GCC extended asm experts please look at this
I _thought_ I was an expert in gcc's extended asm feature, but I can't figure out why this won't compile when optmization is disabled: === #define xchgl(v, m) ({ \ int __result; \ __asm __volatile ("xchgl %0, %1"\ : "=r"(__result), "=m"(m) \ : "0"(v), "1"(m)); \ (__result); }) void lock80386_acquire(volatile int *lock) { while (xchgl(1, *lock) != 0) while (*lock != 0) ; } === It compiles and works fine with -O or higher; but without -O gcc says: locktest.c: In function `lock80386_acquire': locktest.c:11: inconsistent operand constraints in an `asm' This happens with both gcc-2.95.2 (the version in -current) and with the much older gcc-2.7.2.3. I believe the code is correct according to the documentation in the gcc info pages. I tried changing several things anyway to make it more conservative, but I haven't been able to make it compile without optimization. Can any of you see an error in the code? John To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
VPNs and FreeBSD
Has anyone done this yet? I've just acquired this shiny new cable modem and would like to have secure access to my place of work (even though they're only 10 minutes walk away!) Stephen -- The views expressed above are not those of PGS Tensor. "We've heard that a million monkeys at a million keyboards could produce the Complete Works of Shakespeare; now, thanks to the Internet, we know this is not true."Robert Wilensky, University of California To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: UDF (DVD fs)
Soren Schmidt had the audacity to say: Uhm, the real value of UDF is that it can be used as a "real" rw filesystem on CDRW/DVDRAM media, if this is not implemented the value of having UDF is very limited IMHO Not necessarily, since the cd9660 backward compatibility is not a requirement of the standard. I have Pulp Fiction here, and it doesn't work, because it has no 9660 compatibility. In the meanwhile you should be able to mount most modern DVDs using the ISO9660 filesystem as they should be "bridge" format, (in which there is metadata for both types of filesystems). Endeed, makeing a ro UDF filesystem more or less useless :) -Søren -- Coleman Kane President, UC Free O.S. Users Group - http://pohl.ececs.uc.edu To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: loader ??? Awwww... come on someone must know !!!
[EMAIL PROTECTED] wrote: I have FreeBSD 4.0-RELEASE (on i386) What I am trying to do is nfs root boot from the loader (?) I do: ok unload kernel ok load diskless_kernel ok set kernel=diskless_kernel ok boot -r -h Now what I understand is the -r flag will tell the kernel to overide the rootdev and use the "staticly linked device" from when you made the kernel ? Now my kernel (the diskless_kernel ) is a diskless kernel ( with BOOTP, NFS_ROOT etc.. ) compiled in. But when I boot like this is just tries to mount /dev/ad0s1a ... (normal device ) even though it gets the proper reply from the bootp server that its rootfs=nfs mount it even prints : rootfs is my nfs mount Mounting root from ufs:/dev/ad0s1a but still mounts the ufs ! How do if force it to boot root fs on nfs ? I have tried to do a boot -a then when I prompts me for root fs I enter the proper nfs mount and it is cool .. but I need it to do it with out promting... I have tried to : ok set rootdev=nfs:my nfs server ip : mount ok boot -h ok can't determine root device I have also passed it kernel flags ? (mybe wrong ones ... ? ) Now I have looked at the /usr/src/sys/boot/i386/libi386/bootinfo.c and I see where it checks the roodev to see if it is proper . but I cannot determine what type of device I can name so that this will not die on me ??? set rootdev=??? how do I have the kernel to load the nfs root instead of use the currdev ? What does etherboot do when it loads the kernel ? does it pass it special parameters ?? If you are wondering why I just don't boot from etherboot etc...(I do ) but... I want to be able to boot my remote systems without a floppie and chose how to boot from loader (via serail console ...I don't have serial access for the bios but I can get serial console with FreeBSD !! ). Either default (use local drive) or load the diskless kernel , then boot diskless. This way I almost the same options . and maybe automate it... Any help would be great !! thank you kindly nathan To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-net" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: VPNs and FreeBSD
On Sun, 2 Jul 2000, Stephen Hocking wrote: Has anyone done this yet? I've just acquired this shiny new cable modem and would like to have secure access to my place of work (even though they're only 10 minutes walk away!) Well, yeah..ipsec, SSH tunnel, all sorts of SSL-based stuff in ports, SKIP, ... Go for ipsec is my recommendation. See the handbook. Kris -- In God we Trust -- all others must submit an X.509 certificate. -- Charles Forsythe [EMAIL PROTECTED] To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
ls -G seems to depend on TERM=xterm-color
The new improved ncurses version of the color ls stuff doesn't display for me if I am in a "normal" xterm. If I set TERM equal to xterm-color I'm back in business. Is this expected behavior? If so, a note to that effect in the man page would help reduce user confusion. I'll be happy to submit a patch if needed. If it isn't expected behavior, where could I be going wrong? I don't do anything unusual with termcap. Thanks, Doug -- "Live free or die" - State motto of my ancestral homeland, New Hampshire Do YOU Yahoo!? To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: GCC extended asm experts please look at this
John Polstra wrote: I _thought_ I was an expert in gcc's extended asm feature, but I can't figure out why this won't compile when optmization is disabled: = == #define xchgl(v, m) ({ \ int __result; \ __asm __volatile ("xchgl %0, %1"\ : "=r"(__result), "=m"(m) \ : "0"(v), "1"(m)); \ (__result); }) void lock80386_acquire(volatile int *lock) { while (xchgl(1, *lock) != 0) while (*lock != 0) ; } = == It compiles and works fine with -O or higher; but without -O gcc says: locktest.c: In function `lock80386_acquire': locktest.c:11: inconsistent operand constraints in an `asm' This happens with both gcc-2.95.2 (the version in -current) and with the much older gcc-2.7.2.3. I believe the code is correct according to the documentation in the gcc info pages. I tried changing several things anyway to make it more conservative, but I haven't been able to make it compile without optimization. If I change it to use a static inline function, it seems to work and will generate identical code (with -O): static inline int xchgl(int v, volatile int *m) { int __result; __asm __volatile ("xchgl %0, %1" : "=r"(__result), "=m"(*m) : "0"(v), "1"(*m)); return __result; } void lock80386_acquire(volatile int *lock) { while (xchgl(1, lock) != 0) while (*lock != 0) ; } It appears to generate valid code without -O, but I am not 100% sure. It is very inefficient without -O. (Beware the different indirection of "m") Cheers, -Peter -- Peter Wemm - [EMAIL PROTECTED]; [EMAIL PROTECTED]; [EMAIL PROTECTED] "All of this is for nothing if we don't go to the stars" - JMS/B5 To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: ls -G seems to depend on TERM=xterm-color
I'm experiencing the same thing. ls --color doesn't seem to work for me unless like Doug, I set TERM=xterm-color. On Sun, 2 Jul 2000, Doug Barton wrote: The new improved ncurses version of the color ls stuff doesn't display for me if I am in a "normal" xterm. If I set TERM equal to xterm-color I'm back in business. Is this expected behavior? If so, a note to that effect in the man page would help reduce user confusion. I'll be happy to submit a patch if needed. If it isn't expected behavior, where could I be going wrong? I don't do anything unusual with termcap. Thanks, Doug -- "Live free or die" - State motto of my ancestral homeland, New Hampshire Do YOU Yahoo!? To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: GCC extended asm experts please look at this
In article [EMAIL PROTECTED], Peter Wemm [EMAIL PROTECTED] wrote: If I change it to use a static inline function, it seems to work and will generate identical code (with -O): Thanks! I'll give that a try in both places where I'm having this problem. It appears to generate valid code without -O, but I am not 100% sure. It is very inefficient without -O. OK, I'll look at the code closely and try to convince myself that it's correct. I don't care if it's inefficient without optimization. I'm just trying to avoid having to field problem reports from people who for some reason build world without optimization. (This is for the dynamic linker.) Thanks again for the help! John To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Invalidating PACK!!!
On Sat, Jul 01, 2000 at 10:49:11PM +0200, Mustafa N. Deeb wrote: well I think I've eliminated most of these things, I've the disks out of the server and made the cooler point the air to it.. and It does not look like bad cables... about the BAD hd, I've the problem on 5/5 new drives that I bought.. I'll try upgrading the firmeware and BIOS.. and see.. Hi, Where I work we have the same problem we were not however able to get rid of this problem by elimination ie. we replaced scsi cable, scsi adapter and like you we replaced the drive (same specs same seagate) twice with the same results. We also added a fan for the drive noticing it became quite hot with no improvement. da1: SEAGATE ST39236LW 0005 Fixed Direct Access Last that remains is probly a swap of the motherboard which seems wrong since the motherboard doesn't touch the drives at all ie. there is a controller inbetween. Power supply we thought may cause it too but swapping the power cables to drives this did not seem a problem. We've sticked this drive into another host of almost the same configuration and ran bonnie continuously on it as so that we would put continuous use on the drive with no errors reported. Sincerely, Peter Philipp (PP2441) Daemonic Networks "We will survive our loss and we will remember" - RFC 2468 To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Anyone tried StarOffice 5.2 yet?
Hopefully some industrious soul will update the port... Stephen -- The views expressed above are not those of PGS Tensor. "We've heard that a million monkeys at a million keyboards could produce the Complete Works of Shakespeare; now, thanks to the Internet, we know this is not true."Robert Wilensky, University of California To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: mbuf re-write(s), v 0.1
On Thu, 29 Jun 2000, David Greenman wrote: We used to do this in FreeBSD, but found that it was a bad idea for performance reasons. Freeing and reallocating memory from the high-level VM system is quite expensive and the trend in NICs these days is towards needing the code to be even faster, not slower. Further, if the 'peak' is reached often, then you're probably not really gaining much by freeing the memory back to the common pool. What was previously done at some point was use the kernel malloc() to allocate mbufs. As you know, this is a general purpose allocator that has to first determine what algorithm to use and then store the object correctly according to its size. This allocator is faster than that one. This allocator knows that it only has to deal with mbufs and knows that all of these mbufs are of the same size. Yes, malloc is slow for other reasons, but it is especially slow when VM pages are freed back to the general pool. Of course it is possible to introduce hysteresis in the algorithm such that it doesn't free the pages as often, but this (and all the tunables that you proposed) has the negative effect of making the allocator more complex. We've tried very hard not to do this in the current mbuf allocator, making it nearly as efficient as you can get. I guess I just don't see the problem on any of the servers that I manage (ftp.cdrom.com and ftp.freesoftware.com, for example). There are peaks in usage, but they tend to reach the peaks often enough that freeing the pages for short term memory gain is just a waste of CPU cycles. Memory is so cheap these days that throwing memory at the problem seems to be a very reasonable solution, especially when the system clearly needs it during the peaks. -DG David Greenman Co-founder, The FreeBSD Project - http://www.freebsd.org Manufacturer of high-performance Internet servers - http://www.terasolutions.com Pave the road of life with opportunities. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: VPNs and FreeBSD
On Sun, 2 Jul 2000, Stephen Hocking wrote: Has anyone done this yet? I've just acquired this shiny new cable modem and would like to have secure access to my place of work (even though they're only 10 minutes walk away!) I have done just that with nos-tun and Road Runner service. I have not yet implemented the IPSEC feature for security, but the basic tunneling seems to work. Nick Rogness - Speak softly and carry a Gigabit switch. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Invalidating PACK!!!
Thus spake Peter Philipp ([EMAIL PROTECTED]): On Sat, Jul 01, 2000 at 10:49:11PM +0200, Mustafa N. Deeb wrote: well I think I've eliminated most of these things, I've the disks out of the server and made the cooler point the air to it.. and It does not look like bad cables... about the BAD hd, I've the problem on 5/5 new drives that I bought.. I'll try upgrading the firmeware and BIOS.. and see.. Where I work we have the same problem we were not however able to get rid of this problem by elimination ie. we replaced scsi cable, scsi adapter and like you we replaced the drive (same specs same seagate) twice with the same results. We also added a fan for the drive noticing it became quite hot with no improvement. Likewise here. We also changed power supply, to no avail. I had this problem with a 2940UW Pro, with Seagate disks - ST318436LW, and ST318275LW. These are 18Gb Baracuda (7200rpm) units. It was suggested I disable write caching, and this seems to have been the problem. Nick. -- From a Sun Microsystems bug report (#4102680): "Workaround: don't pound on the mouse like a wild monkey." To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: Anyone tried StarOffice 5.2 yet?
I d/l'd from Sun and it installed without a hitch. It is a hell of a lot faster than 5.1 and they've gotten rid of some of the crapisms. Stephen Hocking had the audacity to say: Hopefully some industrious soul will update the port... Stephen -- The views expressed above are not those of PGS Tensor. "We've heard that a million monkeys at a million keyboards could produce the Complete Works of Shakespeare; now, thanks to the Internet, we know this is not true."Robert Wilensky, University of California To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message -- Coleman Kane President, UC Free O.S. Users Group - http://pohl.ececs.uc.edu To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
mbstowcs/wcstombs
hi, there! sorry for posting here but freebsd-i18n list seems to be dead (is it?) some products (e.g. Xerces for C from XML Apache Project) rely on the following behaviour of mbstowcs/wcstombs: when `dest' param is NULL then `len' parameter is ignored and these functions return required length to store the result. This feature is implemented at least in Solaris' libc and glibc 2.1.2. attached patch implements this behaviour (with updated man page and one minor fix in man 3 multibyte). btw is there ongoing effort to merge Citrus libxpg4 to our base tree? freebsd-i18n archives are empty. searching yields nothing except links to Citrus homepage and NetBSD `whatsnew' pages. /fjoe --- lib/libc/locale/ansi.c.orig Mon Jul 3 09:58:32 2000 +++ lib/libc/locale/ansi.c Mon Jul 3 10:05:29 2000 @@ -105,15 +105,21 @@ char const *e; int cnt = 0; - if (!pwcs || !s) + if (!s) return (-1); - while (n-- 0) { - *pwcs = sgetrune(s, MB_LEN_MAX, e); - if (*pwcs == _INVALID_RUNE) + while (!pwcs || n-- 0) { + wchar_t wc; + + wc = sgetrune(s, MB_LEN_MAX, e); + if (pwcs) + *pwcs = wc; + if (wc == _INVALID_RUNE) return (-1); - if (*pwcs++ == 0) + if (wc == 0) break; + if (pwcs) + pwcs++; s = e; ++cnt; } @@ -129,23 +135,30 @@ char *e; int cnt, nb; - if (!pwcs || !s || n INT_MAX) + if (!pwcs || (s n INT_MAX)) return (-1); nb = n; cnt = 0; - while (nb 0) { + while (!s || nb 0) { + char c[MB_LEN_MAX + 1]; + size_t c_cnt; + if (*pwcs == 0) { - *s = 0; + if (s) + *s = 0; break; } - if (!sputrune(*pwcs++, s, nb, e)) + if (!sputrune(*pwcs++, s ? s : c, s ? nb : MB_LEN_MAX, e)) return (-1);/* encoding error */ if (!e) /* too long */ return (cnt); - cnt += e - s; - nb -= e - s; - s = e; + c_cnt = s ? e - s : e - c; + cnt += c_cnt; + if (s) { + nb -= c_cnt; + s += c_cnt; + } } return (cnt); } --- lib/libc/locale/multibyte.3.origMon Jul 3 10:08:00 2000 +++ lib/libc/locale/multibyte.3 Mon Jul 3 10:20:50 2000 @@ -77,7 +77,7 @@ and code each basic element as a sequence of C .Va char Ns s . Individual basic elements may map into one or more -.Pq up to Dv MB_CHAR_MAX +.Pq up to Dv MB_LEN_MAX bytes in a multibyte character. .Pp The current locale @@ -176,6 +176,14 @@ .Fa nwchars wide characters are stored. A terminating null wide character is appended if there is room. +If +.Fa wcstring +is a null pointer, +.Fn mbstowcs +returns the length required to convert the entire array regardless +of the value of +.Fa nwchars , +but no values are stored. .Pp The .Fn wcstombs @@ -189,6 +197,14 @@ .Fa mbstring . Partial multibyte characters at the end of the string are not stored. The multibyte character string is null terminated if there is room. +If +.Fa mbstring +is a null pointer, +.Fn wcstombs +returns the length required to convert the entire array regardless +of the value of +.Fa nbytes , +but no values are stored. .Sh "RETURN VALUES If multibyte characters are not supported in the current locale, all of these functions will return \-1 if characters can be processed, @@ -216,11 +232,15 @@ .Pp The .Fn mbstowcs -function returns the number of wide characters converted, +function returns the number of wide characters converted (or required if +.Fa wcstring +is NULL), not counting any terminating null wide character. The .Fn wcstombs -function returns the number of bytes converted, +function returns the number of bytes converted (or required if +.Fa mbstring +is NULL), not counting any terminating null byte. If any invalid multibyte characters are encountered, both functions return \-1.
Re: mbstowcs/wcstombs
hi, there! On Mon, 3 Jul 2000, Max Khon wrote: sorry for posting here but freebsd-i18n list seems to be dead (is it?) some products (e.g. Xerces for C from XML Apache Project) rely on the following behaviour of mbstowcs/wcstombs: when `dest' param is NULL then `len' parameter is ignored and these functions return required length to store the result. This feature is implemented at least in Solaris' libc and glibc 2.1.2. attached patch implements this behaviour (with updated man page and one minor fix in man 3 multibyte). aargh, there is already opened PR for this (bin/17694) anyway the question is open /fjoe To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message
Re: ls -G seems to depend on TERM=xterm-color
On Sun, 2 Jul 2000, Sean Lutner wrote: I'm experiencing the same thing. ls --color doesn't seem to work for me unless like Doug, I set TERM=xterm-color. That's because the "color" escape sequences are defined for xterm-color in termcap; xterm-color is defined as a superset of xterm (see the tc= directive). I have a lot of X resources defined, and here is one that could help you guys: XTerm*termName: xterm-color Happy to help :) -- Brian Fundakowski Feldman \ FreeBSD: The Power to Serve! / [EMAIL PROTECTED]`--' To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message