ladmins that cannot type a command
good for today and future ladmins that cannot type a command. Any USEFUL proposals that add some real functionality? Since this will enable more people to run FreeBSD that otherwise wouldn't give it a second glance, I would say it is VERY useful. Really? How useful is FreeBSD going to be to someone who cannot type? Who is the target here? Idiots? People without fingers? I see lots of chatter about adding new stuff, much of which is ... rather questionable. Meanwhile many of the features FreeBSD already has are broken. No shortage of PRs that need fixing. Many of these PRs have been sitting around, unloved, for *years*. Some of them even contain patches (that *work*)! If you're building a race car, make sure the engine runs before spending your time polishing the wheels. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: receiving aerial digital TV
[ This topic is better discussed in -multimedia@, suggest that followups drop -hackers@ ] i am out of current knowledge about common TV for about 10 years. Currently in Poland there is aerial TV broadcasted in DVB-T standard. There are TVs with builtin decoder/demodulator or separate decoders/demodulator with HDMI output. But how about just receiving demodulated data and letting mplayer play it, and - most importantly - recording it directly. I know there are USB receivers on market. Are such devices supported under FreeBSD? if so - what driver, what chipset to look when buying such a thing? any other recommendation. With analog TV broadcast there were PCI cards with brooktree chipset that worked with FreeBSD. There are PCI and PCIe tuner cards, USB tuners, and Ethernet tuners. There used to be firewire tuners but these have disappeared the last time I looked. Each type has advantages and disadvantages. Cards require expansion slots, which you may or may not have available. Ethernet tuners allow locating the tuner near the antenna, giving less signal loss in the coax. And they don't need a slot, and are OS independent. Unfortunately, there seems to be only 1 brand of Ethernet tuner (Silicondust HDhomerun). The ATSC/8VSB version has better debugging info than other tuners, but there are many reports that other tuners give better reception. I don't know if this applies to the DVB-T version. Word is that some of the very small USB tuners suffer from poor reception. A high-gain outdoor antenna will give much better reception (due to less multipath) than a low gain antenna. Garbage in garbage out. If you manage to get a signal that is too strong, an attenuator is very inexpensive. Yes, you should be able to record to a file on disk and watch it later. Cron(8) and at(1) are useful for this. Some people like mythtv. Many of the cards use a cx88 family chip. cx88 (in ports) supports many of these cards. http://corona.homeunix.net/cx88wiki Ethernet tuners http://www.silicondust.com/ Useful forums for tuners, antennas, how to fix reception problems, ... http://www.avsforum.com/ Useful info about antennas and reception http://www.hdtvprimer.com ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: HDD write cache
Wojciech writes: after reading quite recent topics about disabling/enabling write cache, i tried to test in on desktop 3.5 drive kern.cam.ada.write_cache: 1 kern.cam.ada.read_ahead: 1 kern.cam.ada.0.read_ahead: -1 kern.cam.ada.0.write_cache: -1 i tried writing 1 or 0 to kern.cam.ada.0.write_cache, and there were exactly no differences at all, and disk seems to do always write caching. Does that drive lie and ignore commands or i do it wrong? this is my disk. ada0 at ahcich0 bus 0 scbus0 target 0 lun 0 ada0: SAMSUNG HD501LJ CR100-10 ATA-8 SATA 2.x device ada0: 300.000MB/s transfers (SATA 2.x, UDMA6, PIO 8192bytes) ada0: Command Queueing enabled ada0: 476940MB (976773168 512 byte sectors: 16H 63S/T 16383C) 1) See the other followups about making sure this change actually takes effect. 2) How, exactly, are you testing this? With Command Queueing enabled, (and assuming that it is working properly) you will not see any difference in speed with a casual test. Does ata(4) support your controller? Last time I looked, ata did not support NCQ. :-( So turning the write cache on/off gives a huge difference in speed. When ahci and siis came out, NCQ was so fast that it looked like the write cache was on. I had to write a special pathological test to be able to tell the difference between having the write cache on and using NCQ. I needed to verify that the write cache was really off so that my data was safe. In anything resembling normal use, NCQ with the write cache off is just as fast as having the write cache on. You really get the best of both worlds, speed and safety. Now all we need is NCQ support for all the controllers that have it. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: IBM blade server abysmal disk write performances
Stefan writes: I seem to remember, that drives of that time required the write cache to be enabled to get any speed-up from tagged commands. This was no risk with SCSI drives, since the cache did not make the drives lye about command completion (i.e. the status for the write was only returned when the cached data had been written to disk, independently of the write cache enable). Interesting. Is there a way to tell, other than coming up with some way to actually test it, whether a particular drive waits until the data has been written to non-volatile memory (the platters in conventional disks) before sending the command completion message? I'm having thoughts of putting sensing resistors in the disk's power cable, attaching an oscilloscope, and displaying the timing of data on the data cable along with power usage from seeking. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: IBM blade server abysmal disk write performances
Wojciech writes: If computer have UPS then write caching is fine. even if FreeBSD crash, disk would write data That is incorrect. A UPS reduces the risk, but does not eliminate it. It is impossible to completely eliminate the risk of having the write cache on. If you care about your data you must turn the disk's write cache off. If you are using the drive in an application where the data does not matter, or can easily be regenerated (e.g. disk duplication, if it fails, just start over), then turning the write cache on for that one drive can be ok. There is a patch that allows turning the write cache on and off on a per drive basis. The patch is for ata(4), but should be possible with other drivers. camcontrol(8) may work for SCSI and SAS drives. I have yet to see a USB-to-*ATA bridge that allows turning the write cache off, so USB disks are useless for most applications. But for most applications, you must have the write cache off, and you need queuing (e.g. TCQ or NCQ) for performance. If you have queuing, there is no need to turn the write cache on. It is inexcusable that FreeBSD defaults to leaving the write cache on for SATA PATA drives. At least the admin can easily fix this by adding hw.ata.wc=0 to /boot/loader.conf. The bigger problem is that FreeBSD does not support queuing on all controllers that support it. Not something that admins can fix, and inexcusable for an OS that claims to care about performance. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: IBM blade server abysmal disk write performances
Scott writes: If I had my way, the WC would be off, everyone would be using SAS, and anyone who enabled SATA WC or complained about I/O slowness would be forced into Siberian salt mines for the remainder of their lives. Actually, If you are running SAS, having SATA WC on or off wouldn't matter, it would be SCSI's WC you'd care about. :-) The bigger problem is that FreeBSD does not support queuing on all controllers that support it. Not something that admins can fix, and inexcusable for an OS that claims to care about performance. You keep saying this, but I'm unclear on what you mean. Can you explain? For most applications you need the write cache to be off. Having the write cache off is fine as long as you have queuing. But with the write cache off, if you don't have queuing, performance sucks. Like getting only 6% of the performance you should be getting. Some of the early SATA controllers didn't have NCQ. Knowing that queuing was very important, I made sure to choose a mainboard with NCQ, giving up other useful features to get it. But FreeBSD does not support NCQ on the nforce4-ultra's SATA controllers. Even the sad joke of an OS Linux has had NCQ on nforce4 since Oct 2006. But Linux is such crap it is unusable. Linux is slowly improving, but I don't expect to live long enough to see it become usable. Seriously. I've tried it several times but I have completely given up on it. Anyway, even after all these years the supposedly performance oriented FreeBSD still does not support NCQ on nforce4, which isn't some obscure chip. they sold a lot them. I've added 3 additional SATA controllers on expansion cards, and FreeBSD supports NCQ on them, so the slow controllers limited by PCIe-x1 have much better write performance than the much faster controllers in the chipset with all the bandwidth they need. I can't add more controllers, there aren't any free slots. The nforce will remain in service for years, aside from the monetary cost, silicon has a huge amount of environmental cost: embedded energy, water, pollution, etc. And there are a lot of them. Wojciech writes: That is incorrect. A UPS reduces the risk, but does not eliminate it. nothing eliminate all risks. Turning the write cache off eliminates the risk of having the write cache on. Yes you can still lose data for other reasons. Backups are still a good idea. But for most applications, you must have the write cache off, and you need queuing (e.g. TCQ or NCQ) for performance. If you have queuing, there is no need to turn the write cache on. did you tested the above claim? i have SATA drives everywhere, all in ahci mode, all with NCQ active. Yes, turn the write cache off and NCQ will give you the performance. As long as you have queuing you can have the best of both worlds. Which is why Karim's problem is so odd. Driver says there is queuing, but performance (1 write per rev) looks exactly like there is no queuing. Maybe there is something else that causes only 1 write per rev but I don't know what that might be. Peter writes: Apart from continuous whinging and whining on mailing lists, what have you done to add support for queuing? Submitted PR, it was closed without being fixed. Looked at code, but Greek to me, even though I have successfully modified a BSD based device driver in the past giving major performance improvement. If I were a C-level exec of a Fortune 500 company I'd just hire some device driver wizard. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: IBM blade server abysmal disk write performances
Matthew writes: There is also no information in the original email as to which direction the I/O was being sent. In one of the followups, Karim reported: # dd if=/dev/zero of=foo count=10 bs=1024000 10+0 records in 10+0 records out 1024 bytes transferred in 19.615134 secs (522046 bytes/sec) 522 KB/s is pathetic. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: IBM blade server abysmal disk write performances
I am thinking that something fancy in that SAS drive is not being handled correctly by the FreeBSD driver. I think so too, and I think the something fancy is tagged command queuing. The driver prints da0: Command Queueing enabled and yet your SAS drive is only getting 1 write per rev, and queuing should get you more than that. Your SATA drive is getting the expected performance, which means that NCQ must be working. Please let me know if there is anything you would like me to run on the BSD 9.1 system to help diagnose this issue? Looking at the mpt driver, a verbose boot may give more info. Looks like you can set a debug device hint, but I don't see any documentation on what to set it to. I think it is time to ask the driver wizards why TCQ isn't working, so I'm cc-ing the authors listed on the mpt man page. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: IBM blade server abysmal disk write performances
Karim writes: dd to the raw drive and no compression/encryption or some other features, just a naive boot off a live 9.1 CD then dd (see below). The following results have been gathered on the FreeBSD 9.1 system: # dd if=/dev/zero of=toto count=100 100+0 records in 100+0 records out 51200 bytes transferred in 1.057507 secs (48416 bytes/sec) By raw drive I meant something like dd if=/dev/zero of=/dev/da0 bs=1m count=1000 of=toto implies that you are using a filesystem. (FFS? ZFS? other?) Matthew writes: But dmesg claims Command Queueing enabled, so you should be getting more than one op per rev, and writes should be fast. Queueing would only help if your load threw multiple ops at the drive before waiting for any of them to complete. I'd expect a dd to a raw device to throw a single, wait for it to return complete, then throw the next, leading to no more than 1 op per rev. I see a huge speedup from NCQ on both raw disks and with FFS/su. Without NCQ I only get 6% of the expected performance, even with a large blocksize. The kernel must be doing write-behind even to a raw disk, otherwise waiting for write(2) to return before issuing the next write would slow it down as Matthew suggests. Writing an entire 3TB disk (raw disk, no fs) gives: 21378.98 real 2.00 user 440.98 sys or 140 MB/s (133 MiB/s) on slow controller in PCIe x1 slot. The same test on the same make model disk on a much faster controller in the chipset takes over 10x as long because FreeBSD does not support NCQ on that controller. :-( Karim's data sure looks like 1 op per rev. Either it isn't really doing NCQ or the filesystem is doing something to keep NCQ from being effective. For example, mounting the fs with the sync option would probably have that effect. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: IBM blade server abysmal disk write performances
I wrote: The kernel must be doing write-behind even to a raw disk, otherwise waiting for write(2) to return before issuing the next write would slow it down as Matthew suggests. And a minute after hitting send, I remembered that FreeBSD does not provide the traditional raw disk devices, e.g. /dev/rda0 with an 'r'. (Now if I could just remember *why* it doesn't.) ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: IBM blade server abysmal disk write performances
25.9 MB/s Even Linux is pretty slow. Transfer rates: outside: 102400 kbytes in 0.685483 sec = 149384 kbytes/sec middle:102400 kbytes in 0.747424 sec = 137004 kbytes/sec inside:102400 kbytes in 1.051036 sec = 97428 kbytes/sec That's more like it. I assume these numbers are reading. You should get numbers nearly this high when writing. Can you try writing to the bare drive without a filesystem? time dd if=/dev/da0 of=/dev/null bs=124k count=25 time (dd if=/dev/zero if=/dev/da0 bs=124k count=25; sync) Between writing more data than the size of memory and the sync, this should hopefully reduce any buffering effects down into the noise and make the numbers more comparable between FreeBSD and Linux. (and more honest) Also eliminates any effect from the filesystem, which will be different between FreeBSD and Linux. Writing should be almost as fast as reading. Is the disk healthy? Smartctl might give a clue. If the disk is healthy and you still get numbers that indicate one write per rev without a filesystem, then the question is why does the driver claim queueing but not deliver it? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: IBM blade server abysmal disk write performances
Karim writes: It is quite obvious that something is awfully slow on SAS drives, whatever it is and regardless of OS comparison. We swapped the SAS drives for SATA and we're seeing much higher speeds. Basically on par with what we were expecting (roughly 300 to 400 times faster then what we see with SAS...). Major clue there! According to wikipedia: Most SAS drives provide tagged command queuing, while most newer SATA drives provide native command queuing [1] Note that the driver says Command Queueing enabled without specifying which. If the driver is trying to use SATA's NCQ but the drive only speaks SCSI's TCQ, that could explain it. Or if the TCQ isn't working for some other reason. See if there are any error messages in dmesg or /var/log. If not, perhaps the driver has extra debugging you could turn on. Get TCQ working and make sure your partitions are aligned on 4 KiB boundaries (in case the drive actually has 4 KiB sectors), and you should get the expected performance. [1] http://en.wikipedia.org/wiki/Serial_attached_SCSI ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Vice versa of 'pkg_info -W'
Domagoj writes: I've attempted to grep '/usr/ports/*/*/pkg-plist' for 'bin/lynx' and shot myself in a foot! :P Even if it did returned sane amount of matches, speed was atrocious. time find /usr/ports/ -name pkg-plist | xargs grep bin/lynx$ /usr/ports/finance/ledgersmb/pkg-plist:@dirrm ledger-smb/bin/lynx /usr/ports/finance/sql-ledger/pkg-plist:@dirrm sql-ledger/bin/lynx /usr/ports/japanese/lynx/pkg-plist:bin/lynx /usr/ports/japanese/lynx-current/pkg-plist:bin/lynx /usr/ports/www/lynx/pkg-plist:bin/lynx /usr/ports/www/lynx-current/pkg-plist:bin/lynx real0m2.065s user0m0.379s sys 0m1.338s Not bad for a system that has been repeatedly insulted on this list recently. (I'm assuming that 6 matches is a sane amount.) ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: FreeBSD for serious performance?
If the driver is doing something daft like DELAY(x) in a fast interrupt handler which would lead to that behaviour, it should be fixed. If it's doing a DELAY(x) in a critical section, it shuld be fixed. They are doing *something* that completely locks out everything else. It is always a device driver. Now, it's quite likely you hit some kind of ata(4) bug which kept it in a tight loop Hard to imagine locking everything out for 19 minutes without being in a loop. So it was likely just spun in some high priority loop that nothing lower-priority could really do anything about. Would several different drivers have this same bug? The next time it happens, please break into the debugger and grab some debugging output. Show alllocks, ps, should be a good couple of things to start with. I've only caught it hanging forever once. It only takes a few milliseconds to cause incoming data to be lost, so I usually don't know about it until looking at the log file later. Not that I could jump into the debugger and gather data in a few milliseconds even if I knew when it was happening. BTW, how do I break into the debugger and gather data when all of the devices are locked out, including the console? I assume that once it recovers, there is no point in gathering data. Alternately - please find a currently actively maintained SATA chipset. The ata controller is soldered to the mainboard, a gazillion pins I'm sure, and no doubt requires very specialized equipment to replace, and I don't know of any pin-compatable replacements. Besides the hardware itself has never caused any problems. The problem is caused by the software, it is the software that needs to be fixed. Ata isn't maintained? Why the bleep not? Disk drivers are essential. I was under the impression that siis(4) and ahci(4) were actively maintained? I'm running four sata controllers using three different drivers and all three drivers lock out other drivers for too long when something unusual happens. And other, non-disk drivers have the same problem of locking out other drivers, even during normal operation. And this happens on yet other drivers on other people's hardware, not just mine. help migrate the nvidia chipset support out of ata(4) I've looked at several of FreeBSD's device drivers (including, as you might expect, ata, siis, and ahci) and I can't make heads or tails out of any of them. Back before FreeBSD existed, I did manage to make a significant improvement to a driver in a BSD-derived system, so I'm not a complete idiot. Several different drivers cause the same problem. Are they all making the same mistake? Or is there a problem in something they all use? Whether a design problem or an implementation bug. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: FreeBSD for serious performance?
Which device drivers? We can't fix problems we don't know about. ata(4) completely hung the system for 19 minutes (at which point I manually intervened, see the PR), probably an infinite loop. http://www.freebsd.org/cgi/query-pr.cgi?pr=170675 Siis(4) and ahci(4) have also caused data loss, presumably by blocking interrupts for too long. Improving these drivers would be wonderful. But better yet, can we please find a way to fix the underlying problem? When a device driver handles an interrupt, it needs to block further interrupts while it modifies its data structures. Otherwise another interrupt coming in might cause it to mangle the data. Right? But! Why does it need to block interrupts for everything? Why does a disk driver need to block interrupts from Ethernet? Why does Ethernet need to block Firewire? Why does Firewire need to block USB? And so on. Can't the disk driver block just its own interrupts and leave the other devices alone? That way, when some device driver writer puts in DELAY(TOO_LONG), at least the other devices will still work. Alternately, why couldn't the data structures be protected with a mutex? Then the drivers shouldn't have to block even themselves. Alternately, why can't drivers have a polling option? Yes, the extra overhead of polling sucks, but losing incoming data sucks a lot more. I am not suggesting that polling should be the default, just an option for those who need it. Alternately, some method I haven't thought of Current machines can have multiple disks, multiple Ethernets, multiple pretty-much-any-device, multiple CPUs, etc. etc. We have SMP kernel to juggle those multiple CPUs. But we still have this absurd bottleneck where the device drivers bring everything to a screaching halt every time an interrupt happens. And if the driver has a bug, or thinks there is a problem and decides to keep DELAY()ing over and over, the entire machine just locks up and stays locked up, often forever. It isn't just me. I have seen quite a few threads where other people are having the same problem. This needs to be fixed. (Fixing this is at *least* a Usenix paper.) ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: NetBSD's boot select MBR
Domagoj writes: MBR supports max of 4 slices/partitions. 4 primary partitions, there are the extended/logical partitions, which allow more. The '/usr/mdec/mbr_bootsel', which you've mentioned, is equivalent to FreeBSD's '/boot/boot0', which is tuned via 'boot0cfg' util. It will also show menu, to choose from, a slice to boot or another drive OR = 'F6' to choose PXE boot. If nothing has been selected, it will simply boot an active slice. So I see FreeBSD's, as superior (additionally supports PXE) = 8.0 onward ... If I understand this correctly FreeBSD's boot0 allows you to set what to boot from the next time and go away while the system reboots. And this method allow booting from not only any slice but any disk, or PXE? But, it does this by writing to the MBR, yes? Some sysadmins want the MBR to be read-only for various reasons. OSes other than FreeBSD aren't going to have boot0cfg. If you're running NetBSD and want to boot Plan 9, do you have to reboot to FreeBSD, run boot0cfg, then reboot to Plan 9? As far as I know, NetBSD's MBR works with any combination of OSes. If the next OS to be booted is on the 1st disk and you don't mind writing to the MBR, you can do the set-and-walk-away thing by setting the active partition. Or you can use something like grub as the next stage and edit grub's default using any text editor, including sed in a script. If you want to boot a slice that is not on the first disk you have to either sit there and type at the menu, or use something like grub. As far as I know, NetBSD's MBR has no support for PXE, but then PXE is intended for diskless machines. I thought NetBSD's had a MBR code, which could be told via a tool, from fully booted and running FreeBSD: Active slice is 2, but please, boot slice 4 at a next boot. After that, I would '# reboot' and without any menu, slice 4 would be booted, for that one time, instaed of an active slice 2. I don't know of a way to do that with just NetBSD's MBR. You could either change the active flag which would remain that way until you changed it again, or sit there and watch for the menu. Given the limited space in the MBR it would be difficult or impossible to add that without giving up something. That feature could be added to a later stage bootstrap. Next time I build a boot disk I'm going to take another shot at multibooting with GPT. I really want to get away from the MBR kludgefest. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
NetBSD's boot select MBR
[ from the FreeBSD for serious performance? thread ] So I use NetBSD's MBR for disks I want to boot from. Can I have a CMD sequence? First would be ... # fetch ... Read NetBSD's fdisk(8) and mbr(8). The MBR is only 512 bytes, and must contain the code and data. This is very limiting, so there are multiple versions depending on which features you need. IIRC, /usr/mdec/mbr_bootsel is the one I use. You can specify a label (up to 7 characters) for each bootable partition (slice in FreeBSD-speak). When the system executes the MBR code, it presents a menu with these labels. You can type '1' for the first partition/slice, '2' for the second, and so on. You can also type a function key to run the MBR of a different disk. The disk numbering can be rather insane, so it must be using the bios disk numbering. If you don't type anything it times out and boots the active partition/slice. Sadly, the FreeBSD bootstrap insists on booting the partition/slice marked active in the MBR. (The MBR code should look at the active bit, the later bootstraps should not.) So if you want to have multiple FreeBSD versions, it works best to put each on its own disk. The last time I used fdisk, large disks triggered a bug where some variable overflowed, giving negative numbers. Hopefully this is fixed by now. FreeBSD's fdisk also had some bug, so I had great fun handcrafting a MBR. It's been a long time since I installed NetBSD from scratch, (I usually just unpack the tar files, run installboot, and edit some config files.) I assume you can just boot a CD and get a shell, and experiment with fdisk on a scratch disk without actually installing NetBSD. To avoid a disaster, do something like dd if=$DISK of=/boot/MBR_backup bs=512 count=1 ; sync beforehand, in case you accidently mess up your boot drive. man pages: http://netbsd.gw.com/cgi-bin/man-cgi?++NetBSD-current isos: http://www.netbsd.org/mirrors/ for example: ftp://iso.netbsd.org/pub/NetBSD/iso/5.2/amd64cd-5.2.iso or ftp://iso.netbsd.org/pub/NetBSD/iso/5.2/i386cd-5.2.iso ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: FreeBSD for serious performance?
Ronald writes: the last Alpha to be produced was shipped way back in 2004... eight years ago... with a top speed of 1.3 GHz I now have a cheap little media player thingy sitting on my desk, and _each_ of its two cores runs faster than tha\ t. In short, Alphas hardly constitute high-end hardware in this day and age. So clock rate is the only thing that matters in your world? Yea, pretty much. As regards to reliability, except for the occasional low-level quirk (which is usually taken care of for me by the kernel guys) I've never had a processo\ r fail on me. Once, about five or six years ago I accidentally burnt up an Athlon XP (by not having the heatsink properly seated) but that was entirely my fault. I care about data integrity, so things like ECC are on my must-have list. I suspect that your cheap little media player thingy doesn't have ECC. If you don't care about getting the correct answer you can have infinite speed. A high clock rate doesn't help when some device driver does block_all_interrupts(); while(1) DELAY(MIGHT_AS_WELL_BE_FOREVER); At least four device drivers have caused me to lose data this way. Not what I call high performance. In my world, high-end means high quality. It doesn't necessarily mean fast, or recent. Data integrity, and yes, reliability, that sort of thing. Performance-wise, in most cases I don't expect to get 99.% of the theoretical best case, I'm usually happy with 90-95%. But without NCQ I'm only getting ~6% of what I should be getting. Pretty pathetic for an OS that claims to be all about performance. All the more so when the crappy OSes do support NCQ on that chip. It's not some rare, obscure chip. Lots of boxes have it. I never found a way to boot from different partitions, much less different disks with GPT. Having just been recently convinced to switch over to GPT (from MBR) I do most sincerly hope that you are either joking or mistaken about this. I am neither joking nor mistaken. I looked but could not find a way. I am not claiming that a way does not exist, merely that I couldn't find it. Perhaps there wasn't a way when I looked (it was awhile ago) but does exist now? I have never been a fan of MBRs, they are for pee-cees with the expected ugly kludges and limitations, real machines don't use them. GPT isn't perfect, but it seems much nicer than MBRs. Warren writes: Grub (or grub2) can do it. Back when I was triple-booting FreeBSD, NetBSD and Linux I used grub (rev number forgotten). It was supposed to be able to boot BSD from a partition but I never got that to work. I had to have it boot the MBR of a different disk which then booted BSD. I wrote 3 little shell scripts that edited grub's menu to change the default. So I could be running FreeBSD and type boot_netbsd, and go have lunch while it rebooted. Other than not booting BSD from a partition it wasn't that bad for something that smells of penguin. If you're booting multiple versions of FreeBSD, see gpart(8) for some partition attributes that may help. You mean the bootme bootonce stuff? That looks promising, assuming I can manage to decode the man page, and figure out what it actually does. Mostly multiple versions of FreeBSD. (for example 7.0 had a couple of bugs preventing it from booting, so having 6 still available was essential.) I no long need Linux (YEA!!!), and after a certain fubar incident have declared a Linux Free Zone. It would be nice if I could also boot Net/Open/Dragonfly. I don't see a way to boot multiple disks, but GPT allows enough partitions that I probably won't care. Not sure if the bootme bootonce stuff wasn't there yet when I looked, or if I just missed it. Thanks for the pointer. Or even consider ZFS boot environments. I plan to stick with FFS w/softdeps. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: FreeBSD for serious performance?
Ronald writes: the last Alpha to be produced was shipped way back in 2004... eight years ago... with a top speed of 1.3 GHz. I now have a cheap little media player thingy sitting on my desk, and _each_ of its two cores runs faster than that. In short, Alphas hardly constitute high-end hardware in this day and age. So clock rate is the only thing that matters in your world? In my world, high-end means high quality. It doesn't necessarily mean fast, or recent. Warren writes: A. Use ahci(4): The ada device driver takes full advantage of NCQ, when supported. Achi and siis support NCQ, but neither attach to the nforce4-ultra, which does support NCQ. I knew that NCQ would be required for acceptable performance and gave up other useful features to get it. Silly me, assuming that the performance orientated version of BSD would support such an essential performance feature on a very popular chipset. Linux has had NCQ on nforce4 since Oct 2006. Without NCQ, writes are slower than USB2, even with purely sequential writes (to the raw drive, no filesystem, so no seeking other then to the next track): device r/s w/s kr/s kw/s wait svc_t %b controller ad6 1726.5 0.0 217535.0 0.0 1 0.6 96 ata3-master ad6 0.0 109.5 0.0 13794.8 1 9.1 100 ata3-master da0 249.9 0.0 15741.4 0.0 1 6.0 100 umass-sim0 ad6 is ST3000DM001-9YN166 on nforce4-ultra chipset da0 is ST3000DM001-9YN166 USB3 disk on a USB2 port (from nforce) Sorry no write performance data for da0, I yanked the USB-to-SATA bridge off after doing some testing. The bridges they're currently shipping are slightly less crappy than they used to be, but they are still crap. B. Use GPT, which does not have the CHS baggage. It is easier and more versatile. My systems with GPT disks don't complain about track alignment. Or maybe that's ahci(4)'s doing. I never found a way to boot from different partitions, much less different disks with GPT. So I use NetBSD's MBR for disks I want to boot from. FreeBSD's and NetBSD's fdisk are both broken. Building MBRs by hand is such fun. Non-boot disks with multiple partitions use GPT. Big data disks get newfs-ed directly, no partitioning. The useless CHS baggage hangs around for decades, but useful hardware loses all support 5 nanoseconds after the last machine is sold. Other useful hardware waits years hoping to get support. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: FreeBSD for serious performance?
[ lack of SATA NCQ support for nforce4-ultra ]
Adrian writes:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux.git;a=commitdiff;h=e2e031eb09760c36099ac127eeb175e06d257aef
which is:
The mcp61 has bug with ncq.
- { PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP61_SATA), SWNCQ },
- { PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP61_SATA2), SWNCQ },
- { PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP61_SATA3), SWNCQ },
+ { PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP61_SATA), GENERIC },
+ { PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP61_SATA2), GENERIC },
+ { PCI_VDEVICE(NVIDIA, PCI_DEVICE_ID_NVIDIA_NFORCE_MCP61_SATA3), GENERIC },
http://en.wikipedia.org/wiki/GeForce_6_Series
says: MCP61 introduced a bug in the SATA NCQ implementation.
As a result, Nvidia employees have contributed code to disable NCQ operations
under Linux.
I have not found a description of the bug.
But FreeBSD says I have:
nVidia nForce CK804 SATA300 controller port
0x9f0-0x9f7,0xbf0-0xbf3,0x970-0x977,0xb70-0xb73,0xcc00-0xcc0f
mem 0xfebfb000-0xfebfbfff irq 21 at device 7.0 on pci0
nVidia nForce CK804 SATA300 controller port
0x9e0-0x9e7,0xbe0-0xbe3,0x960-0x967,0xb60-0xb63,0xb800-0xb80f
mem 0xfebfa000-0xfebfafff irq 22 at device 8.0 on pci0
The CK804 (a few lines higher in the source file) did not get changed to
GENERIC.
As far as I can figure out, the bug (whatever it is) is limited to the MCP61.
___
freebsd-hackers@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
FreeBSD for serious performance? (was: Re: 9.x -- New Install -- serious partition misalignment)
Ronald writes: This probably wouldn't be such a big deal if we were just talking about Linux. But FreeBSD has always prided itself on being a serious OS for serious people with serious work to do... like major server farms and such. In the context of high-end applications on high-end hardware where people are often trying to squeeze out that last drop of performance, Linux is certainly a steaming pile of crap. BSD is orders of magnitude better, but hey, that doesn't take much. But don't brag about high-end hardware. But FreeBSD has dropped support for even semi-high-end hardware (DEC Alpha). So I'm stuck running it on AMD64. Nothing against AMD, they did what they could to try and make a silk purse (amd64) out of a sow's ear (x86). But even getting what passes for a high quality board in amd64/x86 land with good reviews doesn't compare. The firmware is absolute crap, and it's not like it is something you can ignore. BTW, real high end hardware is redundant, better than mil-spec, and provides better than 5-9s uptime. Been there, done that. Several chips/features aren't supported properly. PRs sit for years on end. Performance has been degraded by a whopping 75% ! Having a 4KiB misalignment is nothing compared with not having NCQ support. (Which even linux has, btw.) 25% performance would be a massive upgrade. Or even worse, having the disk driver go into an infinite loop with interrupts blocked, so *nothing* happens and all your incoming data is lost until you manually intervene. Speaking of alignment, I still get partition 1 does not end on a track boundary messages. FreeBSD has no clue where the track boundaries are and neither do I. Disks have used varying numbers of sectors/track for longer than FreeBSD has existed. This is your idea of serious? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: old style kernel configuration
Julian writes: it is however a good way to get mismatching kernel and userland but that's not what we are discussing. The method recommended on http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/kernelconfig-building.html # make buildkernel KERNCONF=MYKERNEL is also a good way to get mismatching kernel and userland. Or any other way of building just the kernel rather than everything. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Memory reserves or lack thereof
Alan writes: In conclusion, I think it's time that we change M_NOWAIT so that it doesn't dig any deeper into the cache/free page queues than M_WAITOK does and reintroduce a M_USE_RESERVE-like flag that says dig deep into the cache/free page queues. The trouble is that we then need to identify all of those places that are implicitly depending on the current behavior of M_NOWAIT also digging deep into the cache/free page queues so that we can add an explicit M_USE_RESERVE. find /usr/src/sys | xargs grep M_NOWAIT | wc -l 2101 Sounds like a lot of work that would need to happen atomically. Would this work: M_NO_WAIT do not sleep, do not dig deep unless M_USE_RESERVE also set M_USE_RESERVE dig deep M_NOWAIT M_NO_WAIT | M_USE_RESERVE (deprecated) New code avoids using M_NOWAIT. Existing code continues working the same way. As time permits, old code is converted to new flags. Eventually M_NOWAIT goes away. Pro: the amount of code that needs to change atomically is much smaller. Con: (1) Have to remember (or look up) difference between M_NOWAIT and M_NO_WAIT. Maybe calling the new flag M_NO_SLEEP would help? (2) Would M_NOWAIT really ever go away? The spl() calls haven't, even after some cage rattling. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: FreeBSD 8.0 suddenly freezing
In the last 72 hours, we've had two different systems freeze; they don't apparently recognize any interrupts, they won't respond to ping, and they require a powercycle to reboot. We can't easily generate an NMI on these boxes. Just on the off chance, does this sound like a familiar -- and preferably soluble -- problem? Only theory I can think of is that some device driver shuts off all interrupts and then goes into an infinite loop. I have had trouble with at least four different device drivers locking everything out long enough to lose incoming data. Usually it recovers without a reboot. Not an infinte loop, just way way too long to leave interrupts shut off. See also the How to diagnose system freezes? thread from July-August. Yuri found that a particular rev of the nvidia driver went into an infinite loop. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: How to diagnose system freezes?
Yuri writes: Anything else I can try? One thing of importance here is that there is an older graphics card 9400 GT on this system and current nvidia-driver-295.71 has an issue with 9400 GT: it makes graphics to malfunction (unpainted windows, long delays switching to terminal mode) or freezes Xorg (but not OS). So I run the older nvidia-driver-285.05.09 which appears to work. That's why I think that nvidia driver is probably to blame for these periodic OS freezes. Also the latest driver version must be, obviously, working for most people because (I think) they mostly have newer than mine nvidia cards. Have you found a way to trigger the bug on demand? Since you suspect the nvidia-driver-285.05.09, try some other driver, and do whatever triggers the bug and see if you get the freeze. So maybe I should also just get the newer nvidia card and shut up, not sure. If you can demonstrate that the various nvidia drivers are broken in various ways, submit a problem report to whoever wrote the drivers (Nvidia presumably). If Nvidia supports their products, then they will fix their drivers. If they don't support their products, why would you want to reward them by buying another nvidia card? But the question I have is: why are device drivers allowed to freeze the entire machine? I have at *least* 4 drivers that freeze the machine long enough for data to be lost. My theory is that: 1) An interrupt comes in. 2) ALL INTERRUPTS ARE BLOCKED ! 3) The device driver sits around too long. 4) Eventually the interrupts are turned back on. If the device driver gets stuck in an infinite loop, the machine hangs forever. Assuming my theory is correct (anyone disagree?), then WHY are ALL interrupts blocked? Why can't we just block interrupts for that particular device? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: How to diagnose system freezes?
Adrian writes: the cycle over a short period may not be the driver writing the same crap to the card. I've seen similar failure modes in windows where during playback, if the system hangs for whatever reason, the card plays the last sample over and over in a loop. Ok, that makes sense. So it could be any device driver. Seems like diagnosing a device driver that is stuck in an infinite loop (or equivalent) with all interrupts shut off would be rather difficult. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: How to diagnose system freezes?
Yuri writes: One of my 9.1-BETA1 systems periodically freezes. If sound was playing, it would usually cycle with a very short period. And system stops being sensitive to keyboard/mouse. Also ping of this system doesn't get a response. So the sound continues, on and off, while everything else freezes? I would suspect that the sound device driver is locking out other devices and hogging the system. WHY does FreeBSD allow one device driver to lock out out everything else? And what can we do about it? I've seen a device driver lock out everything else for minutes-on-end, which of course resulted in data being lost. Losing data is one of the worst things an OS can do. This needs to be fixed. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Awful FreeBSD 9 block IO performance in KVM
da0: 3.300MB/s transfers da0: Command Queueing enabled root@freebsd:/root # dd if=/dev/zero of=/dev/da1 bs=16384 count=262144 4294967296 bytes transferred in 615.840721 secs (6974153 bytes/sec) 1) Does a larger block size (bs=1m) help? 2) That's roughly the speed I'd expect without queueing. Is it really making effective use of queueing, or is something limiting queueing to one transfer at a time? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Pull in upstream before 9.1 code freeze?
Why not create a command wtf(1)? there are really lot of good features that can be made in FreeBSD. actually good, instead of that crap While this is certainly not the most important improvement that could be made (Fix the PRs!), the proposed wtf command could be useful. And, importantly, putting this functionality into a seperate utility is the correct way to do it. We don't want to add more complexity to the shell for several good reasons (bloat, possiblilty of bugs, slowness after every typo, not the Unix way[1], etc.). [1] Make each program do one thing well. To do a new job, build afresh rather than complicate old programs by adding new features. http://www.faqs.org/docs/artu/ch01s06.html ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Training wheels for commandline
As long as it can be toggled off system-wide, persistently (sysctl?), I can't see the harm in bringing that in. It violates the Unix Philosophy. Make each program do one thing well. To do a new job, build afresh rather than complicate old programs by adding new features. http://www.faqs.org/docs/artu/ch01s06.html ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
geom metadata (Re: [CFC/CFT] large changes in the loader(8) code)
These schemes to just put the metadata in some special location and have all the tools know about it create a lot of problems. There is always some tool that doesn't know. There is always some human that doesn't know. Telling the difference between real metadata and some other data that happens to look similar. Convoluted logic that is prone to bugs. I have seen complaints from people that have lost data when some tool wrote metadata on top of it. Losing data is absolutely unacceptable. There is a time to be clever and a time to just keep it simple. Define a FreeBSD geom metadata GPT partition type. Create a 6 sector (3 KiB) FreeBSD geom metadata GPT partition just after the GPT header. PMBR pri GPT header pri GPT table FreeBSD geom metadata data partition(s) sec GPT table sec GPT header Advantages: 1) All OSes will know that this space is taken. 2) Humans looking at the GPT partition table will know that this space is taken, and what it is being used for. 3) The 1st data partition becomes 4 KiB aligned, which is important for many recent disks (yes the metadata partition is not 4K aligned, but is presumably accessed only rarely, so it is not a performance problem) Disadvantages 1) uses up a partition type 2) uses up a partition With GPT neither of these disadvantages is significant. Alternately one could make the geom metadata partition smaller and add a spaceholder partition to get 4K alignment. Yes you can just leave a hole, but putting a partition there labled 4K_alignment makes it obvious why it is there. So, what have I missed? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Freeze when running freebsd-update
Robert writes: 3) the box is responsive to hitting enter at the console (it produces another login: prompt) Getty is in memory and can run. 5) if I try to login to the console, it lets me enter a username then locks up totally, it does not present me with a password: prompt. Login(1) is not in memory, and the kernel cannot read it from disk for some reason. I can get this symptom by writing a large file to a disk on a controller that FreeBSD doesn't support NCQ on. I assume there is a logjam in the buffer cache. Something trivial like reading login in from disk that would normally happen in well under a second can take many minutes. Perhaps geli is causing a similar logjam? Does it hang forever or is it just obscenely slow? If it truely hangs forever it is probably something else. Is there disk activity after it hangs? Can you try it without geli? systat -vmstat might provide a clue. Well, it is geli. I'm unable to reproduce the freeze on the same exact system with everything else the same except for no geli. I'm going to move this thread over to geom, and continue it there. Thanks for your help! It occurs to me that it will need twice as much memory for disk i/o. 1 buffer for encrypted and 1 for unencrypted. I know nothing about geli, so I don't know if it uses the buffer cache for both, or what. Could it be that the kernel isn't keeping enough memory free and manages to paint itself into a corner and not have space to store the unencrypted version of disk reads, and can't page/swap anything out to make space because it doesn't have space to store the encrypted version to write? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Freeze when running freebsd-update
Robert writes: 3) the box is responsive to hitting enter at the console (it produces another login: prompt) Getty is in memory and can run. 5) if I try to login to the console, it lets me enter a username then locks up totally, it does not present me with a password: prompt. Login(1) is not in memory, and the kernel cannot read it from disk for some reason. I can get this symptom by writing a large file to a disk on a controller that FreeBSD doesn't support NCQ on. I assume there is a logjam in the buffer cache. Something trivial like reading login in from disk that would normally happen in well under a second can take many minutes. Perhaps geli is causing a similar logjam? Does it hang forever or is it just obscenely slow? If it truely hangs forever it is probably something else. Is there disk activity after it hangs? Can you try it without geli? systat -vmstat might provide a clue. Well, it is geli. I'm unable to reproduce the freeze on the same exact system with everything else the same except for no geli. I'm going to move this thread over to geom, and continue it there. Thanks for your help! It occurs to me that it will need twice as much memory for disk i/o. 1 buffer for encrypted and 1 for unencrypted. I know nothing about geli, so I don't know if it uses the buffer cache for both, or what. Could it be that the kernel isn't keeping enough memory free and manages to paint itself into a corner and not have space to store the unencrypted version of disk reads, and can't page/swap anything out to make space because it doesn't have space to store the encrypted version to write? I think that's probably about what is happening. I'm still waiting for an answer on the geom mailing list, but I will do some testing with increasing memory sizes and see where the problem stops occurring. Some of the vfs.*buf sysctls might be useful? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Freeze when running freebsd-update
Robert writes: 3) the box is responsive to hitting enter at the console (it produces another login: prompt) Getty is in memory and can run. 5) if I try to login to the console, it lets me enter a username then locks up totally, it does not present me with a password: prompt. Login(1) is not in memory, and the kernel cannot read it from disk for some reason. I can get this symptom by writing a large file to a disk on a controller that FreeBSD doesn't support NCQ on. I assume there is a logjam in the buffer cache. Something trivial like reading login in from disk that would normally happen in well under a second can take many minutes. Perhaps geli is causing a similar logjam? Does it hang forever or is it just obscenely slow? If it truely hangs forever it is probably something else. Is there disk activity after it hangs? Can you try it without geli? systat -vmstat might provide a clue. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: MAGIC with HP KVM - someone will help?
Wojciech writes: 1) i know for a friend that on some motherboards (it was embedded VIA based CPU) geli doesn't work at typing password. 2) in my case (as well as his case) problems happen everytime kernel function cngets is used, it maybe after being unable to mount root in a kernel prompt. The problem IS FreeBSD specific AND some hardware specific. this seems like an interaction between FreeBSD and some hardware (timings?). Really have no idea. I refuse to move from the keyboard I have currently, and occasionally keys get 'stuck' but it's so obvious when it happens it's no big hardship. I have a Belkin adaptor one by the way, and it works but the PROBABLY FReeBSD specific is that some configuration, not only USB to PS/2 adapters, have problems only when kernel cngets routine. my friend switched to linux, in spite of lower performance, on his dedicated fileserver machine with VIA processor, only because keyboard didn't work when geli asked for password at boot, or - for whatever reason - in kernel command prompt activated. it works every other case properly. This reminds me of a problem I have when using a RS-232 console. When the loader is sitting there waiting for a few seconds to see if you will type something, it doesn't always notice keystrokes. I would have to hold the key down (letting it autorepeat) and then it would usually notice. This problem does not happen with the PS/2 keyboard. (no KVM, no USB adapters) If it is a timing problem, perhaps it is related? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: how to turn my computer into a TV
user.vdr writes: As long as there remain some NTSC broadcasts, there might be some that you wish to watch. That's why I wrote: Yes, technically there are still some that exist, for now. However, their death certificate is signed and they're so few that it's not worth mentioning. If you don't think NTSC is worth mentioning, why do you keep posting the same incorrect statements over and over again? You absolutely do NOT have to reencode a stream I did not say anything about RE-encoding anything. Only about encoding/compressing the high bandwidth datastream the tuner generates from NTSC. And to be clear, this only applies to NTSC, not to ATSC. NTSC streams are not broadcast raw. What do you call encoding data that's already encoded if you don't think it's reencoding? Also, doing so causes degredation so unless there's a need for the user to do so, he's better off not wasting his time. NTSC is not a stream of bits. NTSC is analog. The tuner converts the NTSC analog waveform into a raw stream of bits. This raw stream of bits is too large to conviently store on disk, so it needs to be compressed/encoded into mpeg or similar. Some tuners include a hardware encoder, but many do not. Tuners do NOT provide raw audio/video to the system in any case. http://corona.homeunix.net/cx88wiki/Overview/RawVideo While that's technically possible in _some_ cases, and assuming it's fully implemented and functional, I'm unaware of any software that actually provides raw data to the user. I suppose I should have worded my point differently. The cx88wiki URL above describes the cx88 software (in ports). For tuners without a hardware encoder, raw video/audio is the only thing you can get from the tuner when receiving NTSC. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: how to turn my computer into a TV
user.vdr writes: Tuners do NOT provide raw audio/video to the system in any case. http://corona.homeunix.net/cx88wiki/Overview/RawVideo While that's technically possible in _some_ cases, and assuming it's fully implemented and functional, I'm unaware of any software that actually provides raw data to the user. I suppose I should have worded my point differently. The cx88wiki URL above describes the cx88 software (in ports). For tuners without a hardware encoder, raw video/audio is the only thing you can get from the tuner when receiving NTSC. Nope. Prove me wrong. Post the command line to have cx88 (in ports) output encoded (mpeg or similar) video with a pcHDTV HD3000 tuner card receiving a NTSC input. Since you claim that the cx88 output is already encoded, piping the output into mplayer or similar to do the encoding doesn't count. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
(no subject)
user.vdr writes:
Recording doesn't require any compression unless you are transcoding
in real-time. There's no difference between recording ATSC, NTSC, PAL,
etc, and it's actually irrelevant what the stream is.
This is incorrect. ATSC is compressed before broadcast, so
you receive the data already compresed. NTSC and PAL are
broadcast in analog. The tuner performs A-to-D which gives
an uncompressed data stream. Have fun trying to store that.
As a practical matter, you have to compress the data in real time.
Some, not all, tuners include hardware compression.
With very very very few exceptions, all analog NTSC broadcasts have
been switched to digital, by the FCC mandated deadline of June 12,
2009.
As long as there remain some NTSC broadcasts, there might be some
that you wish to watch. That's why I wrote:
You'll need to know if you have any NTSC (analog) stations you
care about or if everything is ATSC (digital).
Aryeh may or may not have any NTSC stations of interest. Given
we are using a broadcast signal only [current US {NYC} standards]
I rather doubt that Aryeh has any PAL stations to worry about.
You absolutely do NOT have to reencode a stream
I did not say anything about RE-encoding anything. Only about
encoding/compressing the high bandwidth datastream the tuner
generates from NTSC. And to be clear, this only applies to
NTSC, not to ATSC.
Tuners do NOT provide raw audio/video to the system in any case.
http://corona.homeunix.net/cx88wiki/Overview/RawVideo
___
freebsd-hackers@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: how to turn my computer into a TV
user.vdr writes: Recording doesn't require any compression unless you are transcoding in real-time. There's no difference between recording ATSC, NTSC, PAL, etc, and it's actually irrelevant what the stream is. This is incorrect. ATSC is compressed before broadcast, so you receive the data already compresed. NTSC and PAL are broadcast in analog. The tuner performs A-to-D which gives an uncompressed data stream. Have fun trying to store that. As a practical matter, you have to compress the data in real time. Some, not all, tuners include hardware compression. Lastly, it's possible to save a single channel or the entire stream which usually contains several channels. Even when saving the full stream, it likely uses far less bandwidth than your media offers so there's no problem there. This appariently refers to ATSC. Yes, modern disks have plenty of bandwidth to store the entire ATSC stream. The main reason to filter PIDs is to save disk *space*. Also, some software can't select which program to decode. Wojciech writes: most people vastly underestimate power of modern CPUs. Many people overestimate the moderness of most people's CPUs. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: EFI development tools
This is a known issue, and had been around for a long time. You can't reliably build 32 bit binaries (what the -m32 flag specifies) on a 64 bit system. The header files (and possibly other things) are wrong. People build 32 bit binaries on 64 bit systems all the time. It is called cross-compiling. I rather doubt that just setting the -m32 flag is sufficient to set up a proper cross-compiling environment. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: how to turn my computer into a TV
[ Added multimedia@ as that is a more appropriate list than hackers ]
I just moved into a very cramped apartment
we are using a broadcast signal only [current US {NYC} standards]
You'll need to know if you have any NTSC (analog) stations you
care about or if everything is ATSC (digital). Hopefully your
building has a good antenna system.
Next, select your tuner(s). You have a choice of tuners that connect
via Ethernet (HDHomeRun, pros: small external box, doesn't need a slot,
works with any OS, better diagnostic info than others. cons: I've seen
a *lot* of postings with people saying that various other tuners get better
reception, digital only no NTSC analog), Firewire (if they are still
available?), USB, PCIe cards (needs a slot), PCI cards (needs a slot).
Some cards also do FM radio.
2. What ports to install
Depends on what tuner(s) you select.
Cards based on cx88 need multimedia/cx88 and multimedia/libtuner
http://corona.homeunix.net/cx88wiki
has a list of supported cards.
Some cards are supported by bktr(4).
Make sure the tuner you select is supported by FreeBSD.
Recording ATSC takes very little CPU. Recording NTSC takes either
a lot of CPU or hardware compression. Decoding either takes a lot of CPU
(or hardware decoding which AFAIK FreeBSD doesn't have). You can use
at(1) for automated recordings. A full ATSC channel is 19.3 Mbps.
Some tuners allow filtering by PID, which saves disk space.
For playback you can use mplayer or any of several similar programs.
___
freebsd-hackers@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: FreeBSD Boot Times
Brandon writes: Booting into Ubuntu minimal or my own custom Linux distro, literally takes 0.5-2 seconds to boot up to shell 0.5-2 seconds from power-on to a shell prompt? How do you get through the firmware that fast, much less firmware plus an OS? Which reminds me, back when I was triple-booting Free, Net, and penguinix, I noticed that when rebooting, penguinix didn't go through all the firmware stuff like the BSDs do. That is one way to save a lot of time, at least for reboots. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Solving the great resource problem
Spending resources to create more releases is pointless when the PRs aren't getting fixed. Oh, Look! Release 9.2.2.2.2.2 is out! The system still crashes every 5 seconds, but a typo on the true(1) man page is fixed. We need a more global discussion about all the things that resources are spent on, and which are the most useful. Replacing perfectly good components simply because they are GPL. The purpose of BSD is supposed to be creating a great OS, not providing software hoarders with a supply of free code to abuse. Sending people to conferences. Nice, but clearly a luxury. Meanwhile the hardware support is a disaster. PRs sit for years and years and years. The documentation has plenty of room for improvement. It seems there are never enough resources to fix problems, but somehow there are always resources to do yet another fork. FreeBSD, NetBSD, OpenBSD, DragonflyBSD, and now Bitrig. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: GSoC Project: Automated Kernel Crash Reporting System - Discussion
1) tar up files 2) encrypt tarball 3) copy encrypted tarball with rcp, ftp, uucp, ... ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: csh builtin command problems
Robert writes: I want this: # echo test\ttest test # cat test test test I have given up on using echo for anything the least bit fancy, in favor of printf(1) which gives much better control. printf test\ttest\n ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Ways to promote FreeBSD?
*WHY* is Linux so much more popular than the BSDs? GPL vs BSDL ? (Create a GPLed BSD and see if it takes off.) the obese cartoon penguin? Do most people actually prefer the lower quality product? Popularity is inversly proportional to quality in many areas, not just OSes. marketing? Is there something that Linux does better than the BSDs, that many people care about? -- Is there a list somewhere of companies/organizations that contribute to FreeBSD? A list of those that freeload (Use FreeBSD as part of their product but do not contribute back.)? Do we need to recrute a diplomatic fund raiser person to shake the bushes? -- Are the resources we do have being used as effectively as possible? Looks to me like funds are being spent on things that would likely happen anyway (because they are fun, would result in a paper, or are commercially useful), while things that need doing but aren't always fun (such as fixing PRs) don't get funding and never get done. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Is there any modern alternative to pstack?
Konstantin Belousov wrote: My opinion is that such tool should be imported into the base. Why? Don't optional tools belong in ports? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
mlock/mlockall (was: Re: problems with mmap() and disk caching)
Andrey writes: Wired memory: kernel memory and yes, application may get wired memory through mlock()/mlockall(), but I haven't seen any real application which calls mlock(). Apps with real time considerations may need to lock memory to prevent having to wait for page/swap. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Graphical Terminal Environment
Brandon writes: I'm still avidly trying to work on this idea, but right now the issue seems to be with AMD and NVIDIA not documenting their protocols. Intel does a good job, but I don't have any Intel chips with graphics laying around. I thought that AMD had documented most of it by now, with the major exception of the UVD? I'm working on a minimal GPU right now on an FPGA. Currently it looks like I could run 4 monitors at 1080p for about $50 Have FPGA prices come down that much? The OGP-D1 was quite a bit more than that, last time I looked. Or would that be the price for a production version with an ASIC? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: mlock(2) man page errata
mlock(2) says: A single process can mlock() the minimum of a system-wide ``wired pages'' limit and the per-process RLIMIT_MEMLOCK resource limit. Shouldn't this say maximum rather than minimum? I don't think so. The minimum of the two would be the limit that you will hit first, and presumably is the point at which you cannot mlock any more pages. Ok, but can mlock() the minimum of is easy to misread as can mlock() at least. Perhaps it would be more clear to say something like The amount of memory that a process can mlock() is limited by the per-process RLIMIT_MEMLOCK resource limit, and by a system-wide ``wired pages'' limit. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Please help me diagnose this crazy VMWare/FreeBSD 8.x crash
Subsequent inspection suggested that it was happening during the periodic daily, though we never managed to get it to happen by manually forcing periodic daily, so that's only a theory. Perhaps due to a bunch of VMs all running periodic daily at the same time? We had a perfectly functional, nearly zero-traffic VM, since Jabber traffic averages no more than a few messages per hour. It was working for quite some time. We moved it from a local datastore to an iSCSI datastore that ended up getting periodically crushed by the load (in particular during the periodic daily load imposed by a bunch of VM's all running at once). At this point, this one VM started hanging on I/O. We expected that this would clear up upon return to a host with a local datastore. It did not. This ended up as a broken VM, one that would hang up overnite, maybe not every night, but several times a week at least. ... For the problem to follow the VM in this manner, and afflict *only* the one VM, strongly suggests that it is something that is contained within the VM files that constitute this VM. That is consistent with the observation that the problem arose at a point where the VM is known to have had all those files moved from one location to a dodgy location. That's why I believe the evidence points to corruption of some sort. Compare a backup of the VM before it broke to a backup of the same VM after it broke. Hopefully the haystack of insignificant differences isn't too large, or the significant difference needle might be a lot of fun to find. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
mlock(2) man page errata
mlock(2) says: A single process can mlock() the minimum of a system-wide ``wired pages'' limit and the per-process RLIMIT_MEMLOCK resource limit. Shouldn't this say maximum rather than minimum? [EAGAIN] Locking the indicated range would exceed either the system or per-process limit for locked memory. [ENOMEM] Some portion of the indicated address range is not allocated. There was an error faulting/mapping a page. In some cases, the code returns ENOMEM if the limit is exceeded. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Please help me diagnose this crazy VMWare/FreeBSD 8.x crash
FreeBSD ?? - 7.4 never crash FreeBSD 8.0 - 8.2 crashes Obvious short term workaround is to run production on 7.4 (assuming you can) until you figure out what is wrong with 8.x. What filesystem(s) are you running? UFS? ZFS? other? started randomly disconnecting people every morning Due to timeouts? Something might be keeping interrupts disabled too long. there were other good reasons to reload the VM, so I nuked the VM, which, of course, fixed it. I can look at recovering the faulty VM from backup Sounds like corruption. Can you compare the bad VM against a good one? If you find corruption, the question then becomes what is causing the corruption? Sounds like the same thing is getting corrupted every time, rather than something at random. Sounds like the corruption is causing a deadlock in something common, like the buffer cache, or filesystem, or... Is it possible to have root be a ramdisk? This might give you access to the utilities, depending on where the problem is. I have vague memories that the sticky bit used to lock a program in memory, but sticky(8) indicates that this is no longer the case. Is there a way to lock a program in memory? (So that it will be available when the system can't do disk i/o.) If not, you could keep some windows open with things like top and systat -vmstat running. Some of the utilities have options to look at a disk file rather than the live system, if you can get a core dump (swap to NFS?). ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Reigning in disk cache buffer hogs
amd64 4 GiB nVidia nForce CK804 (aka nforce4-ultra), SiI3132, JMB363 sata disks FreeBSD 8.2 FFS/SU Problem: I/O to one disk can fill up the disk buffer cache, starving other processes of disk I/O. If the other processes are logging real time data (from a closed source black box), then data is lost. Each process has its own private disk to minimise latency due to seeking and such, but that isn't sufficient to prevent the problem. It is always the onboard nforce4-ultra controller that causes the problem. The expansion card controllers never cause the problem. The onboard nforce4-ultra controller is supported by ata(4). The expansion card controllers are supported by ahci(4) and siis(4). The onboard nforce4-ultra controller is much faster than the expansion card controllers, except that ata doesn't support NCQ, so the nforce has a serious bottleneck writing. Is there something about NCQ other than the obvious performance improvement that would explain why the NCQ controllers never have the problem? What sort of miracle would it take to get NCQ support for the nforce4-ultra controller? Linix has had NCQ support since 2006. Is there some knob to control reserving a few buffers per disk to prevent one disk from hogging all the buffer cache? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Graphical Terminal Environment
Brandon writes: (If you haven't noticed, I'm going to keep finding excuses to write this as I really am kindof excited to learn/work on it) ideas: Display PostScript rio (from Plan 9) If you're mainly looking for a low-level graphics project, maybe reverse engineer something on the GPU (e.g. UVD) ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: OS support for fault tolerance
The problem then is how to feed both machines the same inputs, and compare the outputs. Do we need a third machine to supervise? Can we have each machine keep an eye on the other, avoiding the need for a third machine? A pair would work as long as the only failures are obvious (e.g. crashes). If they simply disagree as to the result, how would we determine which one was right? Depends on what sort of work the machine is doing. If the job is something that can be done again, you could simply try again, if you still get different answers try a third machine or wade in and start manually inspecting things until you find the problem. If the job is time critical or you can't get the same inputs again, then the machine needs to get it right the first time. How many 9s of reliability do you need and how many resources can you throw at it? 2x hardware can be good for better than 5 9s. (high quality hardware and software, and technicians standing by with cold spares) I've heard that mil gear uses 3x hardware. Building a 5 9s system is... non-trivial. So I'm wondering what sort of reliability we can get with 2x off the shelf commodity hardware and a bit of software? Similar to mirroring/RAID but with whole computers rather than just disks. Classic Unix technique of doing 10-20% of the work and getting 80-90% of the result. Which then leads to the issue of how to avoid problems when *it* breaks. For some reason, this reminds me of a Dr. Seuss story: http://www.goodreads.com/review/show/49519038 *grin* Gotta love Dr. Seuss. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: OS support for fault tolerance
Rayson writes: The question is, are we planning to handle 95% of the errors for 99% of the hardware we run on, or are we really planning to spend years trying to design something that would require special hardware support? I assume this started as: Oh look, most CPUs have multiple cores these days, maybe we could play with fault tolerance. Which could be useful if CPU cores failed a lot, but in reality what fails is disks, disks, controllers, disks, random other things, and disks. Assuming you have avoided the garbage-quality stuff, and have the system on a UPS. If you have enough ports you can add more disks and mirror or some other version of RAID. The next step is to duplicate everything. Not by looking for a mainboard with redundant everything, but by simply adding another computer. And rather than getting two of the same machine, you're better off if they are different, so that they don't have the same bugs. The problem then is how to feed both machines the same inputs, and compare the outputs. Do we need a third machine to supervise? Which then leads to the issue of how to avoid problems when *it* breaks. Can we have each machine keep an eye on the other, avoiding the need for a third machine? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Getting PRs fixed
Igor writes: You mean something like: http://people.freebsd.org/~edwin/gnats/ ? Daniel writes: http://www.oook.cz/bsd/prstats/ Yes, something like these. Stephen writes: You should get extra points for difficult PR's. One way to measure this would be to give more points for fixing older PR's than newer PR's. Older might be harder, or it might mean boring, or seen as less important. It might be worth giving more points for old PRs regardless, to help get the old ones fixed. The main goal here is to get PRs fixed. It just feels wrong to have PRs sitting around for years on end, and if it is a significant problem you can be sure that the submitter is unhappy about not having a fix. A longer PR might be more difficult. A complex problem takes longer to describe, and longer to fix, than just there is a typo in the foo(1) man page. One problem is PRs that are closed without being fixed. Some of these are legitimate (dups, submitter error, already fixed in newer release, ...) but some shouldn't have been closed. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Giant lock gone?
The original goal for 5.0 was to completely remove the Giant lock (and do other cool SMP-related stuff). Eventually it was realized that this was too big a goal to fully accomplish in 5.0 (albeit too late in the process) and the goal was changed to do the basic framework for the new SMP model; and lay the groundwork for some things run under Giant for now, and we'll remove it from them ASAP. That actually turned out to last through 6, making 7 the realization of what 5.0 was supposed to be. So you are saying that the Giant lock was completely removed in 7.0? 8.2 says: atkbd0: [GIANT-LOCKED] psm0: [GIANT-LOCKED] Yeah, I think nitpicking minor details about Giant still being used on non-critical drivers is definitely the key to solving the problems that face FreeBSD today. I could have more thoroughly clarified the reality of the when/where/how of Giant but it really wasn't central to my point. Sigh. I wasn't intending to nitpick, I was confused and curious by the apparient contradiction between completely remove and what I saw. 8.2 again: find /usr/src/sys | xargs grep -i giant | wc -l 2018 Skimming though the output of grep, a few are just false positives, but most appear to involve the Giant lock. Perhaps these are all unimportant, I don't know. Apologies to anyone I upset. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: FreeBSD has serious problems with focus, longevity, and lifecycle
John writes: - EOL 7 - mark 8 as legacy - mark 9 as the _only_ production release - release 10.0 in January 2017 Until a few days ago 8 was the latest, shinest release. So you want to suddenly demote it all the way down to legacy? I thought the goal was to have releases that can be used for a long time? On the one hand, many users want/need releases to have a long lifetime. On the other hand, we want to be able to start a new release when some new major feature is mature enough. If these features come out often enough, we end up with a lot of releases to support, and supporting a lot of releases uses up a lot of time and effort. Assuming that a lot more resources aren't going to magically appear, perhaps the solution is to ramp down the level of support for older releases. 7 - legacy (only gets fixes for security, panic/hang, data loss) 8 - supported (security, panic/hang, data loss bugs, plus others as requested) 9 - very supported (most improvements that aren't massively disruptive) stable - not quite bleeding edge, not a release yet, not recommended for production, brave users can try out new features current - bleeding edge The legacy level shouldn't have many fixes, so it shouldn't take too much time and effort. It should be possible to support multiple legacy releases. If fixes only get backported to supported releases when users request it, the amount of time and effort may be low enough? I don't have a good idea of how many requests would be made. If the requests are infrequent enough, it might be possible to support more than one supported release. (Better names for supported and very supported would be welcome.) A time based schedule can make sense if there are a lot of small changes. It doesn't work as well with fewer large changes. If the clock says it is time for a new major release but none of the major new features are ready, it doesn't make sense to do a major release. A time based schedule might make more sense for minor and point releases? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Giant lock gone? (was: Re: ...focus, longevity, and lifecycle)
The original goal for 5.0 was to completely remove the Giant lock (and do other cool SMP-related stuff). Eventually it was realized that this was too big a goal to fully accomplish in 5.0 (albeit too late in the process) and the goal was changed to do the basic framework for the new SMP model; and lay the groundwork for some things run under Giant for now, and we'll remove it from them ASAP. That actually turned out to last through 6, making 7 the realization of what 5.0 was supposed to be. So you are saying that the Giant lock was completely removed in 7.0? 8.2 says: atkbd0: [GIANT-LOCKED] psm0: [GIANT-LOCKED] ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Getting PRs fixed (was: Re: ...focus, longevity, and lifecycle)
Andriy writes: And dealing with PRs is not always exciting. Neither is brushing your teeth or cleaning the kitchen, but most of us manage to do them at least occasionally. Part of being a grown up. Instead of looking for a stick to hold over developers to get them to fix PRs, let's look for carrots to make fixing PRs more appealing. Idea 1: Fix 'n' PRs, get a tee-shirt, fridge magnet, plush daemon, ... Idea 2: Give it status. Set up a web page with PR fixing stats name/handle..total PRs fixed...fixed in last 12 months...average fixed/year Sheldon..150...9072 Leonard..131..11067 Howard...104...2052 Raj...80...8080 ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: FreeBSD has serious problems with focus, longevity, and lifecycle
Atom writes: i bought myself a LENOVO T510 when it first came out, around early 2010. it's got an i5 CPU and Arrandale GPU. it's two years old and on freeBSD i STILL can't run xorg properly with it. I have a machine from 2005-08 that FreeBSD still doesn't support properly in 2012. After much research I figured out that SATA NCQ was an essential feature, and choose a mainboard with nforce4 to get it. FreeBSD still doesn't support NCQ on nforce after all these years. Linux has had NCQ on nforce4 since Oct 2006. FreeBSD also doesn't properly support the onboard VIA firewire chip, which is still found on new mainboards today. I don't necessarily expect support for every exotic chip out there the first day they hit the street, but these are both popular chips, and it is 6.5 years later. I'm not sure how an OS is supposed to have The power to serve when it can't even talk to disks properly. And all the device drivers that think it is ok to lollygag around for absurd lengths of time with interrupts turned off, thus causing data to be lost. Damien writes: Check this PR I opened some months ago: http://www.freebsd.org/cgi/query-pr.cgi?pr=161123cat=kern It was planned for 9.0-RELEASE, there is no mention of 8.x That's just the kind of problem John raises here. Hey, at least you have a fix, and it is even in a release (I'm assuming it made it into 9.0). The bigger problem is all the bugs that don't have fixes at all. Igor writes: patches go ignored (no, I don't have a reference) Here is a PR that contains a patch, yet is still open after several years. Also fixes an even older PR from Dec 2006. Dinky little patch, works great. Should be easy enough to code inspect, test, and check in. http://www.FreeBSD.org/cgi/query-pr.cgi?pr=127717cat= Mark writes: Why is everyone so afraid of running -STABLE? Experience. John writes: Is three per year an insane schedule ? Remember, I am simultaneously advocating that FreeBSD stop publishing two production releases simultaneously, which is almost oxymoronic Why is publishing two production releases almost oxymoronic? Seems like a very good policy to me. Say you are running 8.1. It is good to have the choice of going to a low risk 8.2 with bugfixes or going to 9.0 with some major new feature at the expense of more work and higher risk. If you want the option of sticking with a major release series (say 8.x) for a long time, then there needs to be at least two production releases supported. As fast as major releases are coming out, probably 3 or 4. Why are major releases coming out so often? Gotta compete in the large number war. NetBSD was at 1.x for years and years, then suddenly someone decided to change the numbering scheme and they're off to the races. Firefox has caught the same insanity. I see complaints from medium-to-large sites, yet FreeBSD's budget is so small. Surely it must be possible for these medium-to-large sites to pay into a fund to improve things. FreeBSD clearly needs more developers to fix problems, to code review, test, and check in patches, and to improve the documentation. Judging from this email thread there is a demand for more/better release engineering and backporting as well. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Sound on the system briefly interrupts when kde4 switches windows with nvidia card
I have kde4 on 8.2 with translucency windows effect enabled and nvidia graphics driver. Every time I switch an active window or maximize some window, sound played by mplayer interrupts for ~0.5 sec. Very annoying effect. Problem disappears when windows effects in kde4 are turned off. Is this likely a bug in nvidia driver that it locks up the system when a lot of OpenGL operations are in progress? Or what might be a problem? Some drivers do things like DELAY(big number) while interrupts are turned off. Very bad. Look through the source, maybe you can find it. We REALLY need a way to service one device without shutting off interrupts from other devices. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: cron(8) mis-feature with @reboot long after system startup
The system doesn't go multiuser until the rc jobs complete, even if you attempt to background them with ''. This can be a problem with long running jobs. I started using cron @reboot for this reason. I haven't run into the problem since I've never needed to run /etc/rc.d/cron restart. Add an option to cron to check lastlog and if within 5 or 10 minutes of the last reboot, then call run_reboot_jobs(). Depending on timestamps might be okay as a temporary quick-and-dirty workaround, but there is likely to be a case where it will also do the wrong thing. What if you need to restart cron within the 5-10 minutes? Maybe something like: rc script sets a flag, cron_reboot script checks and resets flag. The flag could be a file ( /tmp/rebooting_system). Better yet, a run-the-reboot-script command line option could be added to cron. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: easy way to determine if a stream or fd is seekable
something like the following inside lseek() would take care of tape drives:
if (S_ISCHR(sb.st_mode) || S_ISBLK(sb.st_mode)) {
if (ioctl(io-fd, FIODTYPE, type) == -1)
err(1, %s, io-name);
if (type D_TAPE)
return(EBADF)
}
I'd suggest ENODEV (Operation not supported by device) rather than
EBADF (Bad file descriptor).
To do this correctly, we'd need some standard way to ask the
device driver if the device can perform the seek or not.
___
freebsd-hackers@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: easy way to determine if a stream or fd is seekable
lseek() on a tape drive does not return an error, nor does it actually do anything. IIRC some tape drives can seek, while others cannot. Vague memories that it is supposed to be possible to put a filesystem on a DECtape and mount the filesystem. It might be that FreeBSD doesn't currently support seeking on a tape, but we shouldn't paint ourselves into a corner by assuming that it is fundimentally impossible. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: checksum offload
The data sheet for intel 82576 advertises IP TX/RX checksum offload but the driver does not set CSUM_IP in ifp-if_hwassist. Does this mean that driver (and chip) do not support IP TX checksum offload or the support for TX is not yet included in the driver? The first question is is checksum offload a good idea? There is less protection against errors. This assumes firmware flaws, right? Albeit, it also does confuse some software like tcpdump. Unfortunately lots of high performance people use checksum offload which removes much of the end to end protection Alan Cox http://kerneltrap.org/Linux/Data_Errors_During_Drive_Communication Bugs in the hardware/firmware/software are not required to have a problem, but if present would make things even worse. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: checksum offload
The data sheet for intel 82576 advertises IP TX/RX checksum offload but the driver does not set CSUM_IP in ifp-if_hwassist. Does this mean that driver (and chip) do not support IP TX checksum offload or the support for TX is not yet included in the driver? The first question is is checksum offload a good idea? There is less protection against errors. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: excessive use of gettimeofday(2) and other syscalls
Firefox 5 and 6 has more gettimeofday call than 2 per second on my amd64-8.2-stable box. i don't see why chromium needs to call gettimeofday(2) or any library function that triggers it more than 3000 times a second. What the BLEEP are web browsers doing that they need the clock so often? I suspect the answer is the same as why firefox uses significant amounts of CPU when it should be idle, why it uses memory without bound (I actually had to add ulimit to my shell's rc file :-( ), and so on. Using links -g, ktrace -di -tc -p6951; sleep 1; ktrace -C; kdump|wc -l gives a typical count of 300-400, highest count seen: 1454. What we need, is a sanely written web browser that has the features we need. Unfortunately the last time I checked, links and dillo both lacked features needed for online shopping/banking. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
optimum fs param sizes (was: ZFS installs on HD with 4k physical...)
I guess formatting the filesystem for 4k sectors on a 512b drive would still work but it would be suboptimal. What would the performance penalty be in reality? It would be suboptimal but only for the slight waste of space that would have otherwise been reclaimed if the block or fragment size remained 512 or 2K. This waste of space is insignificant for the vast majority of users and there are no performance penalties, so it seems that switching to 4K sectors by default for all file systems would actually be a good idea. If you have large files, then large block/frag sizes waste less space than small ones. Remember that keeping track of all those blocks and frags has overhead, and that overhead can waste more space than you save at the end of files. Is anyone looking at extending the range of tuning variables for FFS? Allowing even larger block/frag sizes would be useful, as well as larger cylinder groups and fewer inodes. Fsck runs a *lot* faster with fewer inodes. The expected average file size and expected average number of files per directory limits could use work as well. newfs -e 1 -b 16384 -f 2048 -g 67108864 -h 16 -i 67108864 -U -o space -L 58data /dev/ada9p1 density reduced from 67108864 to 3676160 /dev/ada9p1: 2861588.5MB (5860533100 sectors) block size 16384, fragment size 2048 using 12754 cylinder groups of 224.38MB, 14360 blks, 64 inodes. with soft updates A cylinder group size of 224.38MB is just insanely small with today's disk and file sizes. IIRC the block size limit is 64K, but there was some bug with using block/frag larger than 16K/2K. Even 64K is too small. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: CONF class of files
CONFDIR is for base, not ports
Perhaps ${BASE_CONF_DIR}, ${PORTS_CONF_DIR}, ...
___
freebsd-hackers@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Testing a change to printf(9)
I've been working on fixing problems with printf(9), log(9) and related functions. Today I tried converting printf(9) to write to the log rather than directly to the console, unless the log is not open, in which case the message is also sent to the console. Printf(...) is now the same as log(LOG_INFO, ...). I commented out the line in /etc/syslog.conf that sends some log messages to the console. In multiuser mode, normal printfs go to log, but not the console, as expected. Bootup messages, shutdown messages, and panic messages all show up on the console as expected. Are there any other special cases to test? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Testing a change to printf(9)
I've been working on fixing problems with printf(9), log(9) and related functions. Today I tried converting printf(9) to write to the log rather than directly to the console, unless the log is not open, in which case the message is also sent to the console. Printf(...) is now the same as log(LOG_INFO, ...). oh please no! from my perspective, I want my printf output to go to the console. immediately, regardless of where it goes after that. I don't care if there is or is not a log. I do NOT want to EVER have the problem I've had on linux where the last vital bit of output never made it out before the system stopped. once it's been shown on the console I don't care what happens to it.. Printfs to the console cause data loss. Easily repeatable. Absolutely unacceptable. You are welcome to fix the actual underlying problem. I would love to see the underlying problem fixed! I've asked a few times before, but I'll ask again. Why does a driver for one piece of hardware have to block interrupts for all hardware? I thought changing from spl to mutex was supposed to fix this? My changes do not prevent a sysadmin from having printf output go to the console, it gives them a choice. Currently there is no choice. I commented out the line in /etc/syslog.conf that sends some log messages to the console. In multiuser mode, normal printfs go to log, but not the console, as expected. Bootup messages, shutdown messages, and panic messages all show up on the console as expected. Are there any other special cases to test? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: compiler warnings (was: Re: [rfc] a few kern.mk and bsd.sys.mk related changes)
please keep in mind that -Wfoo does reflect the ideas of the GNU people regarding *proper* code. the warnings themselves are sometimes wrong, because they complain about perfectly correct code. I attempted to get the gcc people to improve this: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=9072 Most of the warnings I see are either due to someone thinking all the world is ILP32 and doing things like storing a 64 bit pointer or long in a 32 bit int, or due to the compiler needing more info to insure that they are not trying to stuff 64 bits into 32, such as missing prototypes. Either way it needs to be fixed. In many cases the developers that claim to write such great code, and claim that the compiler warnings don't matter are the ones whose code has the most bugs (seg faults, floating point exceptions, ...). Pretty much the entire kernel is compiled with quite a large number of warning classes enabled, and -Werror set, for example. Whoever did this, THANK YOU!!! fixing warnings in other people's code is useful only if you can get them to accept the fixes back Fixing bugs is always useful. Certainly it is a *lot* more efficient if you can get them fixed at the source rather than having to maintain patches. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Active slice, only for a next boot
Peter writes: A better approach is to be able to boot whatever slice you want without having to change the active slice. NetBSD can do this. The MBR puts up a menu of the bootable slices on the disk being booted. You can allow the timer to time out and boot the default. Or you can enter the number of the slice you want to boot. Or you can type a function key F1 F2 ... to boot a different disk, and it will load the MBR from that disk and run it. There is an alternative for keyboards without function keys. So can FreeBSD - though only for MBR If so, the documentation could use improvement. Examples would be great. - this functionality doesn't seem to have made it into the GPT bootcode. Is anyone working on this? MBR is only good for 2 TB and the 3 TB disks are becoming cost competitive. I've switched over to GPT for everything except boot disks, as it is less brain damaged than MBR. And it works great. Except that one of the 27 stages of boot code that FreeBSD uses INSISTS on booting the active slice, so you can tell the MBR to boot slice 3 and slice 3's boot code sees that slice 4 is active and boots slice 4. Multibooting worked correctly when I last used it (a few years ago). Have you raised this as a PR? No, partly because I haven't had much luck with PRs. Mainly because I'd rather spend my time trying to migrate to GPT than improving MBR. So many bugs/misfeatures so little time, RS-232 console + hardware modem + POTS = remote console And even that doesn't fully work unless you have a serial-aware BIOS. Real computers have RS-232 consoles. If you neglected to specify RS-232 console in the requirements, there is this thing. I haven't tried it. http://www.realweasel.com/ Somebody probably sells a KVM-over-IP box. You could see if openbios supports your mainboard. John writes: And it works great. Except that one of the 27 stages of boot code that FreeBSD uses INSISTS on booting the active slice, so you can tell the MBR to boot slice 3 and slice 3's boot code sees that slice 4 is active and boots slice 4. There are only 3 stages, It feels like more. :-) and boot1.S is what looks at the active slice. Unfortunately it doesn't have a better way to do this as the only input it gets from boot0 or any other MBR boot loader is the BIOS drive number in %dl. I'm not sure how else you would detect that a non-active slice was booted from when that is your only input. The NetBSD boot code manages to do it. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: Active slice, only for a next boot
I have i.e; 3 slices, of which first is active. Now I wana set slice 2 active, but only for a one/next boot. Once slice 2 is booted and system is shutdown or rebooted, once again, first slice is active and booted, without user's intervention. I think that setting the active slice is the wrong approach. Because then whatever OS you boot must then somehow know to automagically change the active slice back to the default. A better approach is to be able to boot whatever slice you want without having to change the active slice. NetBSD can do this. The MBR puts up a menu of the bootable slices on the disk being booted. You can allow the timer to time out and boot the default. Or you can enter the number of the slice you want to boot. Or you can type a function key F1 F2 ... to boot a different disk, and it will load the MBR from that disk and run it. There is an alternative for keyboards without function keys. And it works great. Except that one of the 27 stages of boot code that FreeBSD uses INSISTS on booting the active slice, so you can tell the MBR to boot slice 3 and slice 3's boot code sees that slice 4 is active and boots slice 4. My ugly workaround is to have my backup FreeBSD slice on a different disk. The better solution would be to fix FreeBSD's boot code so that it boots the slice it is in rather than insisting on booting the active slice. And import the NetBSD boot selector MBR. Anyone has any idea, for a case of a remote server, which is accessible over ssh, only when it is up? Remote access only when the system is up will bite you, sooner or later. The classic formula is: RS-232 console + hardware modem + POTS = remote console ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: compiler warnings (was: Re: [rfc] a few kern.mk and bsd.sys.mk related changes)
maybe we find some nice -Wwarning options which are reasonable to have -Wmissing-declarations -Wimplicit FreeBSD's gcc doesn't seem to have -Wcoercion ??? Bugzilla indicates that it was added years ago (2006?). It would be really really nice if -static worked on (nearly) everything. and - most importantly - don't break tinderbox That's a matter of fixing the bugs before adding the warning flags to tinderbox. Ports need attention. The warnings I get there are frightening. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: compiler warnings (was: Re: [rfc] a few kern.mk and bsd.sys.mk related changes)
Chris writes: Ports need attention. The warnings I get there are frightening. I find it comforting that they're just that: warnings. How do they frighten you? High quality code does not have any warnings. The most frightening thing is the attitute that They're just warnings, so I'll ignore them. Most compiler warnings should be fatal errors. And a lot of the warnings that require a -Wwhatever should be on by default. Code that doesn't compile cleanly often has other problems, like assuming that all CPUs are ILP32 little endian, like not checking return codes, etc. But hey, the next time the weather service issues a tornado warning, feel free to go outside and fly a kite. After all it's just a warning. a13x writes: -Wcoercion seems to have only been a SoC project in 2006 [1]. i checked gcc trunk and it's not in the gcc(1) manual. [1] http://gcc.gnu.org/wiki/Wcoercion And yet someone marked the bug fixed. http://gcc.gnu.org/bugzilla/show_bug.cgi?id=9072 ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: [RELEASE] New Boot-Loader Menu
There's really only room for one or two more menu items. Perhaps some items could be moved to a 2nd level menu? 1) boot multiuser mode ( default ) 2) boot single user mode 3) menu to set boot options 4) help Would be nice: a fix for having to lean on a key autorepeating for a couple seconds. Could you explain? I don't follow. On my Tyan Tomcat k8e 2865, just entering the number rarely if ever works. I have to either repeatedly bang away at the key or hold it down, letting the RS-232 terminal do the autorepeat thing, while hoping that it notices before the timer runs out. Is that with the current code that's in CVS? I'd love for you to try my code on that hardware. One of the things that I worked on in the very beginning was the responsiveness. That's with 6.x, 7,x and 8.x, currently running 8.2. I haven't looked to see if anything has changed in CVS since 8.2. I'll try your code, but I can't promise when. I'm swamped at the moment and windows when the machine is free keep getting smaller and less frequent. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: [RELEASE] New Boot-Loader Menu
___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: [RELEASE] New Boot-Loader Menu
[ attempt #2 - grumble - sorry about the blank message, hope it works this time - grumble- ] I hope that works for serial console. VT100 may be a reasonable default in that case, but it would be good to make sure that menu works even on a dumb terminal. Perhaps we should put 'key' letter in brackets then? This needs to work, correctly, everywhere. This needs to be easy to understand by a stressed out user whose machine is having problems. Therefore: Thou shalt not assume graphics. Thou shalt not assume color. Thou shalt not assume VT100 or any specific terminal. Thou shalt not assume ability to display bold. Thou shalt not assume ability to underline text. Thou shalt not assume availability of multiple fonts. Thou shalt not assume more than 24x80 chars. Thou shalt not assume scrollback. Thou shalt not assume fancy cursor movements. Thou shalt not assume presence of function keys. Thou shalt not assume presence of arrow keys. Thou shalt not assume a fast interface. Thou shalt not assume the three-finger-salute works. Putting brackets around letters (and numbers) sounds good. If there is room, perhaps add a message explaining that the user should enter one of the choices in brackets. A help option would be useful, giving a reminder of what things like ACPI and APIC stand for, what safe mode does, etc. Would be nice: uname -v of the kernel it will boot. Would be nice: a user friendly way to boot from a different disk/partition/kernel. Without the user having to know the mapping between what the firmware calls disks and what FreeBSD calls disks. And without writing anything to disk. Would be nice: a fix for having to lean on a key autorepeating for a couple seconds. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: [RELEASE] New Boot-Loader Menu
Already on the to-do list is to support ``loader_logo=...'' in /boot/loader.conf Including an option for no logo? (For consoles that are slow and/or small, and for people that just don't like the logos.) Putting brackets around letters (and numbers) sounds good. If there is room, perhaps add a message explaining that the user should enter one of the choices in brackets. I think I'm going to have to play with this and see what we come up with. I don't want to make it too busy if you know what I mean. That's with respect to the brackets. As for adding a messages... things are a bit tight and again, I'm afraid of making it too cluttered. I'll post some screenshots of some mock-ups tomorrow, incorporating the various requests. A help option would be useful, giving a reminder of what things like ACPI and APIC stand for, what safe mode does, etc. This is not altogether infeasible. Since this menu (unlike the current one) has the ability to be wiped from screen and then recalled completely in the original state is was left in ... implementing an F1 feature that reads text from a file would be very do-able. If there is a help option that the user can figure out how to execute, the explaination about brackets (if you go that route), entering numbers, letters, and such could be included in the help screen(s) instead of the main menu page. Would be nice: a fix for having to lean on a key autorepeating for a couple seconds. Could you explain? I don't follow. On my Tyan Tomcat k8e 2865, just entering the number rarely if ever works. I have to either repeatedly bang away at the key or hold it down, letting the RS-232 terminal do the autorepeat thing, while hoping that it notices before the timer runs out. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to freebsd-hackers-unsubscr...@freebsd.org
Re: bge dropping packets issue
I'm investigating an issue we are seeing with 6.1-RELEASE and the bge driver dropping packets sporadically at 100MBps speed. Its get mainly aggravated when heavy disk I/O occurs Has anyone seen this problem before with bge? Am I barking up the wrong tree with my initial investigation? Does anyone know if its even possible to achieve 100% packet capture with bge at its supported speeds (10/100/1000)? I had a similar problem with bge and 6.0-RELEASE. Bge works great for me in 6.2-RELEASE. So far 7.0-RELEASE looks good as well (bge-wise, I do have unresolved issues with 7). My app is TCP, cranking the TCP receive buffer way up helped, as did turning off Nagle. My bge is 1000, but connected at 100 since that is what the other end is. I saw problems at less than 20 Mbps. There is still a problem that other drivers can lock out bge for too long. For example kern/118093: firewire bus reset. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: boot loader
You also can't dual boot without console access. You can, I have a system triple booting without using the console. Method 1: boot through grub. Create shell scripts that edit grub's menu.lst file. Method 2: use fdisk(8) to change active partition. With or without dual booting, if the least little thing goes wrong you will need the console, and maybe the reset button, and maybe the hard power switch to recover. Even rebooting without changing anything can be dangerous since fsck might decide to be unhappy. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: SII3512 rev0 ?
Any input on the reliability of SII3512 is also welcomed.
dmesg | grep 3512
satalink0: Silicon Image SATALink 3512 (rev. 0x01)
No reliability problems from the 3512. (up 74 days)
This is on an Alpha box running NetBSD. I haven't tried the 3512
with FreeBSD.
I took a quick look at the drivers, and found a difference.
NetBSD:
dev/pci/satalink.c
/*
* Rev. = 0x01 of the 3112 have a bug that can cause data
* corruption if DMA transfers cross an 8K boundary. This is
* apparently hard to tickle, but we'll go ahead and play it
* safe.
*/
if (PCI_REVISION(pa-pa_class) = 0x01) {
sc-sc_dma_maxsegsz = 8192;
sc-sc_dma_boundary = 8192;
FreeBSD:
dev/ata/ata-chipset.c
if ((ctlr-chip-cfg2 SIIBUG) ch-dma) {
/* work around errata in early chips */
ch-dma-boundary = 16 * DEV_BSIZE;
ch-dma-segsize = 15 * DEV_BSIZE;
}
And of course there may be other differences I didn't find.
___
freebsd-hackers@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: SII3512 rev0 ?
pcirev
{ ATA_SII3512, 0x02, SIIMEMIO, 0, ATA_SA150, SiI 3512 },
{ ATA_SII3512, 0x00, SIIMEMIO, SIIBUG,ATA_SA150, SiI 3512 },
What happened to rev 1?
Any input on the reliability of SII3512 is also welcomed.
dmesg | grep 3512
satalink0: Silicon Image SATALink 3512 (rev. 0x01)
No reliability problems from the 3512. (up 74 days)
This is on an Alpha box running NetBSD. I haven't tried the 3512
with FreeBSD.
___
freebsd-hackers@freebsd.org mailing list
http://lists.freebsd.org/mailman/listinfo/freebsd-hackers
To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Fwd: AMD deciding _now_ what to do about Linux
So who, exactly, is the best person to write to requesting docs for AMD/ATI graphics/video chips? We need to politely inform them that There are a lot of operating systems out there, and they all need high quality, fully functional drivers. FreeBSD, NetBSD, OpenBSD, Plan-9, Open-Solaris, ... We might want to add a ATI graphics/video card to a computer with a different CPU architecture (Alpha, Sparc, PPC, ...) Binary drivers are useless, because: Too many OSes. Too many CPU archs. Can't fix bugs. Can't fix security holes. Therefore we need documentation on how to program the chips. 2D is not enough. We need video decoding and 3D. If we can't have high quality, fully functional source code drivers for the OS and CPU arch of our choice, then there is no reason to buy the product. Have I left out anything? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: sil3124 sata
Just wondering if there is any planned support for this card (or similar) I've added sos@ to Cc list, who may have interest to this as well. Note that developing drivers requires that the developer has his hands on actual hardware and hardware specifications. Exactly, get me the HW on my lab table and I'll do the driver as time/docs permits, its that simple :) -Soren Is getting a sil3124 board on Soren's lab table a job for the FreeBSD Foundation? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: kern/106343: Need SATA NCQ support
In message [EMAIL PROTECTED], Remko Lodder writes: Synopsis: Need SATA NCQ support State-Changed-From-To: open-closed State-Changed-By: remko State-Changed-When: Mon Dec 4 20:29:54 UTC 2006 State-Changed-Why: Hello, this is not a PRoblem but a request for assistance. Please reask this on the -hackers mailinglist (findable via http://lists.freebsd.org/mailman/listinfo). They might be able to help you quicker and sooner then via a PR (Which is not really a PR yet). If after discussion people think you should file a PR, notify me and I will reopen it with all the recent information involved. Anyway: Thanks for taking the time to report this issue and for using FreeBSD http://www.freebsd.org/cgi/query-pr.cgi?pr=106343 I asked on -questions a year ago. No response. IMO the man page(s) should include things like how to turn SATA NCQ on/off. And if the device drivers don't provide the functionality, they should. So IMO this is a PR. But we can run it past -hackers, that's fine. So, for those of you on -hackers just tuning in, can someone please tell me how to turn SATA NCQ on? Thanks! Here's the PR in question: atacontrol cap ad4 reports: Feature Support Enable Value Vendor Native Command Queuing (NCQ) yes - 31/0x1F I have looked in the atacontrol(8) man page, the ata(4) man page, I grepped all the section 4 8 man pages, and I did an online search, but I cannot find how to turn on SATA NCQ. Write performance without NCQ is unacceptable. NCQ would improve write performance significantly. It is not obvious if this is just a documentation problem, or if NCQ support needs to be added to the various SATA controller device drivers. (nforce4 ultra in my case) atapci1: nVidia nForce4 SATA150 controller port 0x9f0-0x9f7,0xbf0-0xbf3,0x970-0x977,0xb70-0xb73,0xcc00-0xcc0f mem 0xfebfb0\ 00-0xfebfbfff irq 10 at device 7.0 on pci0 ata2: ATA channel 0 on atapci1 ad4: 238475MB Seagate ST3250823AS 3.03 at ata2-master SATA150 ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to [EMAIL PROTECTED]
