Re: Is this an SSD problem or a controller problem?
The analysis done by Jeremy is probably right. The device return errors in response to commands disabling capability that it reported as suppoted and enabled. That is not fatal, but just annoying. Send me please output of the 'camcontrol identify ada15 -v' to check. In case of smartctl I guess there is some more problem causing command timeout and following device reinitialization with the same errors. That may already be fixed in 9-stable branch. 15.02.2013 3:13 пользователь Jeremy Chadwick j...@koitsu.org написал: (Please keep me CC'd as I am not subscribed to this list) (Also CC'ing mav@ since he can shed some light on this too) Re: http://lists.freebsd.org/pipermail/freebsd-questions/2013-February/249183.html It is neither an SSD problem nor a controller problem. FreeBSD is issuing a specific ATA CDB command to the SSD, and the SSD rejects this request, returning ABRT status. This is perfectly normal per ATA specification; the error is harmless. You should open a PR on this matter, as FreeBSD should be adjusted in some manner to deal with this situation, either via appropriate workarounds or a drive quirk. mav@ would know what's best. You will need to provide output from the following commands in your PR: * dmesg * camcontrol identify ada15 * pciconf -lvcb * Same lines you did in your Email Further technical details, which you can put into the PR if you want: Looking at src/sys/cam/ata/ata_all.c we can see that the output of the ACB is in bytes, output per ata_cmd_string(). Thus: ACB: ef 90 00 00 00 40 00 00 00 00 02 00 Decoding per T13/2015-D rev 3 (ATA8-ACS2) working draft spec: 0xef = command = SET FEATURES 0x90 = features = Disable use of SATA feature 0x00 0x00 0x00 = lba_*= n/a 0x40 = device = n/a 0x00 0x00 0x00 = lba_*_exp= n/a 0x00 = features_exp = n/a 0x02 = sector_count = Enable/Disable DMA Setup FIS Auto-Activate Optimisation 0x00 = sector_count_exp = n/a DMA Setup FIS is defined as: 7.50.16.3 Enable/Disable DMA Setup FIS Auto-Activate Optimization A Count field value of 02h is used to enable or disable DMA Setup FIS Auto-Activate optimization. See SATA 2.6 for more information. The enable/disable state for the auto-activate optimization shall be preserved across software reset. The enable/disable state for the auto-activate optimization shall be reset to its default state upon COMRESET. This feature has to do with NCQ capability for certain types of DMA transfers. src/sys/cam/ata/ata_xpt.c contains the responsible code. I could be wrong here (mav@ please correct me), but in probestart(), there is: 452 case PROBE_SETDMAAA: 453 cam_fill_ataio(ataio, 454 1, 455 probedone, 456 CAM_DIR_NONE, 457 0, 458 NULL, 459 0, 460 30*1000); 461 ata_28bit_cmd(ataio, ATA_SETFEATURES, 462 (softc-caps CTS_SATA_CAPS_H_DMAAA) ? 0x10 : 0x90, 463 0, 0x02); 464 break; CTS_SATA_CAPS_H_DMAAA is defined per include/cam/cam_ccb.h as Auto-activation, and its name implies DMA, so this would match the feature in question. This would explain why you see it when the machine boots (xpt(4) probe), as well as when smartctl is run or smartd starts (uses xpt(4)). However, I noticed this piece of code in probedone(): 739 /* 740 * Some HP SATA disks report supported DMA Auto-Activation, 741 * but return ABORT on attempt to enable it. 742 */ 743 } else if (softc-action == PROBE_SETDMAAA 744 status == CAM_ATA_STATUS_ERROR) { 745 goto noerror; Which makes me scratch my head -- the comment and logic seems to imply there shouldn't be any error condition reported, but you do see one. This also implies that the drive advertises per SATA protocol DMA AA yet when xpt(4) tries to disable it the drive rejects that request with ABRT. I don't know why OCZ rejects disabling that feature, but whatever. Addendum note for mav@ -- we also need to add an ADA_Q_4K quirk entry to ata_da.c for Vertex 4 SSDs (OCZ-VERTEX4). -- | Jeremy Chadwick j...@koitsu.org | | UNIX Systems Administratorhttp://jdc.koitsu.org/ | | Mountain View, CA, US| | Making life hard for others since 1977. PGP 4BD6C0CB | ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: No sound from speaker, using Realtek ALC269 and snd_hda
On 31.10.2012 03:00, Big Yuuta wrote: Yes, actually I'm using sysctl hw.snd.verbose=4 to understand what's happening inside. From my tests, it's not a sense redirection problem. Because, when I unmute everything, and then I plug a headphone, the sound goes to the headphone and the internal speaker is mutted, and when I unplug it, the internal speaker gets the sound, and (nid 26: the headphone) is mutted. So, anyway, I'm still debugging it and I found out that I don't have to unmute everything, I just have to make sure that nid=15 is never mutted. i.e. the test I wrote in my last email could be more precise like this: hdaa_audio_ctl_amp_set_internal(struct hdaa_devinfo *devinfo, nid_t nid, int index, int lmute, int rmute, int left, int right, int dir) { uint16_t v = 0; if(nid == 15){ //just don't mute nid15, and it works lmute = 0; rmute = 0; } Strange thing is from my dmesg here: http://dpaste.com/818967/plain/ I thought that nid 15 wasn't used. hdaa0: nid: 15 [DISABLED] hdaa0:Name: audio mixer hdaa0: Widget cap: 0x0020010a hdaa0: Input amp: 0x8000 hdaa0: mute=1 step=0 size=0 offset=0 hdaa0: connections: 2 hdaa0: | hdaa0: + [DISABLED] - nid=2 [audio output] hdaa0: + [DISABLED] - nid=11 [audio mixer] Maybe it's disabled because others were disabled for other reasons? I honestly don't grok the whole thing yet. The driver reports as disabled all parts of the CODEC that are unused in specific configuration. It is quite usual to have half of CODEC unused. To avoid unexpected effects driver mutes all disabled controls. According to information reported by CODEC, this mixer is really unused. I see no problem from the driver side there. We, IMVHO, probably just should add a patch inside hdaa_patches.c for this case: case HDA_CODEC_ALC269: if (subid == 0x10438437){ //0x10438437 is my subsystem id. w = hdaa_widget_get(devinfo, 15); if(w != NULL) //some magic to unmute it ? } break; What do you think? I think Realtek engineers got crazy. They not only created several different CODECs sharing the same ID (my laptop also uses variant of ALC269, but it has no such problem), but also violated their own specs and information reported by CODEC. Patch below should hide problematic muter from the driver. Solution is far from perfect, but that is best I can propose without having more information. Please test it and report about results. --- hdaa_patches.c (revision 242352) +++ hdaa_patches.c (working copy) @@ -541,6 +541,21 @@ hdaa_patch(struct hdaa_devinfo *devinfo) if (w != NULL) w-connsenable[0] = 0; break; + case HDA_CODEC_ALC269: + /* +* ASUS EeePC 1001px has strange variant of ALC269 CODEC, +* that mutes speaker if unused mixer at NID 15 is muted. +* Probably CODEC incorrectly reports internal connections. +* Hide that muter from the driver. There are several CODECs +* sharing this ID and I have not enough information about +* them to implement more universal solution. +*/ + if (subid == 0x10438437) { + w = hdaa_widget_get(devinfo, 15); + if (w != NULL) + w-param.inamp_cap = 0; + } + break; case HDA_CODEC_CX20582: case HDA_CODEC_CX20583: case HDA_CODEC_CX20584: -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: No sound from speaker, using Realtek ALC269 and snd_hda
On 31.10.2012 18:24, Big Yuuta wrote: No, that patch didn't do the trick for the simple reason that my actual subid was 0x84371043 and not 0x1043837 -you spot my mistake? ;-) So here's the one that finally got me sound with my Asus EeePC 1001px on the internal speaker: Now I can have a good nap thinking that somehow I contributed a tiny little bit to my favorite OS. Thank you so much, Alexander! Not only you helped me solve this problem that I had for several months, but you also got me interested in FreeBSD's internal code! Thank you for your contribution, it is really valuable, as there is no other way to handle such kind of hardware issues. I've just committed the patch to the HEAD branch and will merge it down to 8/9-STABLE in two weeks. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: No sound from speaker, using Realtek ALC269 and snd_hda
On 30.10.2012 09:17, Big Yuuta wrote: Hi Alexander, I got the sound out of that speaker!! :) I actually started to read your code, and I'm still trying to understand the stuff in it (I never wrote a device driver) Anyway, so I did a little hack'ish modification (just to test) in function hdaa_audio_ctl_amp_set_internal(struct hdaa_devinfo *devinfo, nid_t nid, int index, int lmute, int rmute, int left, int right, int dir) { uint16_t v = 0; // Do not mute, even if asked for. Test to be removed of course lmute = 0; rmute = 0; I know, this is absolutely NOT the way to do it, but I wanted to see if the speaker wasn't muted (or the mixer that controls it) and it -actually- WAS muted! Now, I guess I'll have to read the whole code, try to understand its internals, and I hope to be able to write a patch to add to hdaa_patches.c Speaker should be muted on headphones connection. That is one of two ways of handling playback redirection. But it should be unmuted on disconnection. You may try to set sysctl hw.snd.verbose=4 and connect/disconnect headphones. It should report which controls are affected and how. You may try this hack to use pin controls instead of muters for redirection: --- hdaa.c (revision 242315) +++ hdaa.c (working copy) @@ -260,7 +260,7 @@ /* (Un)Mute headphone pin. */ ctl = hdaa_audio_ctl_amp_get(devinfo, w-nid, HDAA_CTL_IN, -1, 1); - if (ctl != NULL ctl-mute) { + if (ctl != NULL ctl-mute 0) { /* If pin has muter - use it. */ val = connected ? 0 : 1; if (val != ctl-forcemute) { @@ -290,7 +290,7 @@ hdaa_hpredir_handler(struct hdaa_widget *w) continue; ctl = hdaa_audio_ctl_amp_get(devinfo, as-pins[j], HDAA_CTL_IN, -1, 1); - if (ctl != NULL ctl-mute) { + if (ctl != NULL ctl-mute 0) { /* If pin has muter - use it. */ val = connected ? 1 : 0; if (val == ctl-forcemute) -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: No sound from speaker, using Realtek ALC269 and snd_hda
On 28.10.2012 17:12, Big Yuuta wrote: I have an asus eeepc 1001px, on which I installed 9-stable. I, since day one, never had any sound coming out of the speaker, but when I plug a headphone in the jack, I have the sound -in the headphones. The chipset is a Realtek ALC269 which some people said is supported. Now driver support most of CODECs. I don't know cases when specific CODEC would be a problem, and Realtek CODECs from functional side are better then many others. But many systems have custom wiring and laptops are usually more problematic there. I read, and re-read the man page for snd_hda so many times, but I just can't make it work. Here's what I have right now in my device.hints Code: hint.hdaa.0.nid20.config=as=1 seq=0 device=Speaker hint.hdaa.0.nid26.config=as=1 seq=15 device=Headphones hint.hdaa.0.nid18.config=as=2 seq=0 As I see, these three hints are replicating existing configuration and effectively useless. #hint.hdaa.0.gpio_config=0=set hint.hdaa.0.config=forcestereo,ivref50,ivref80,ivref100,ivref,vref hint.pcm.0.config=gpio0,gpio1,gpio2,gpio3,gpio4,gpio5,gpio6,gpio7 hint.pcm.0.vol=100 This is my n'th attempt, I pretty tried everything I could imagine. Here's my verbose dmesg http://dpaste.com/818967/plain/ Am I missing something? Is my understanding of pinouts correct? CODEC configuration looks good and I see no problems in driver output. I think most likely problem is in CODEC wiring and power amplifier. Your CODEC has two GPIO lines and EAPD line. That gives 8 possible combinations. I would recommend you to try them all. GPIOs, as you tried could be set with hint.hdaa.0.gpio_config tunable. EAPD line can be controlled (0 or 100) by the ogain mixer control. Unluckily with this output I can't completely identify your system to check what Linux does for it. Could you send me `devinfo -vr` output. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: No sound from speaker, using Realtek ALC269 and snd_hda
On 28.10.2012 22:09, Big Yuuta wrote: CODEC configuration looks good and I see no problems in driver output. I think most likely problem is in CODEC wiring and power amplifier. Your CODEC has two GPIO lines and EAPD line. That gives 8 possible combinations. I would recommend you to try them all. GPIOs, as you tried could be set with hint.hdaa.0.gpio_config tunable. EAPD line can be controlled (0 or 100) by the ogain mixer control. You mean combinations like: hint.hdaa.0.gpio_config=0=set 1=set 2=set As I've told, there are only two GPIO pins, so only 0=X 1=Y, plus ogain mixer control. Where the values could be set or keep? But then again, the man also cites clear, disable, input. So that'd make more than 8 possible permutations? Or is it something else? Can you please give me an example? GPIO pins are bidirectional, but I have doubt that input or high impedance (disable) state could be used to control something. keep is also useless, as by default these pins are disabled. So you have only set and clear. Also, can I, after booting, just try with: kenv hint.hdaa.0.gpio_config=0=set 1=set 2=set kdunload snd_hda.ko kdload snd_hda.ko Yes, that should work. or should I reboot each time? I don't think it is required. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: No sound from speaker, using Realtek ALC269 and snd_hda
On 28.10.2012 23:20, Big Yuuta wrote: Unfortunately, trying with the different combinations didn't work. I wrote this tiny script just to make sure I'm not forgetting a case: #!/bin/sh echo testing with: $1 - $2; kenv hint.hdaa.0.gpio_config=0=$1 1=$2; kldunload snd_hda.ko; kldload snd_hda.ko; mplayer song.mp3; and I run it like this: ./test.sh set set ./test.sh set clear ./test.sh clear set ./test.sh clear clear And nothing came out of the speaker, alas :( Alas. I'll just remind one more time about `mixer ogain`. Also check that pin sensing is working. Try to plug in/out headphones. With verbose messages enabled, you should see messages about that on console and in logs. What's strange is that the speaker actually works with OSS from ports, but when I use it (OSS) with mplayer it crashes the whole system whenever I try to skip in a video/audio file On Sun, Oct 28, 2012 at 9:58 PM, Big Yuuta init...@gmail.com wrote: On Sun, Oct 28, 2012 at 9:42 PM, Alexander Motin m...@freebsd.org wrote: On 28.10.2012 22:09, Big Yuuta wrote: CODEC configuration looks good and I see no problems in driver output. I think most likely problem is in CODEC wiring and power amplifier. Your CODEC has two GPIO lines and EAPD line. That gives 8 possible combinations. I would recommend you to try them all. GPIOs, as you tried could be set with hint.hdaa.0.gpio_config tunable. EAPD line can be controlled (0 or 100) by the ogain mixer control. You mean combinations like: hint.hdaa.0.gpio_config=0=set 1=set 2=set As I've told, there are only two GPIO pins, so only 0=X 1=Y, plus ogain mixer control. So that would be: hint.hdaa.0.gpio_config=0=set 1=set hint.hdaa.0.gpio_config=0=set 1=clear hint.hdaa.0.gpio_config=0=clear 1=set hint.hdaa.0.gpio_config=0=clear 1=clear which I try with: mixer ogain 0:0 mixer ogain 100:100 I'm going to check once again, but I think that unfortunately that didn't work. I hope that I forgot a combination! Thanks again, Alexander -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: How to enable NCQ on freebsd 8.1 ?
Hi. Nickolay Krylov wrote: I have enabled options ATA_CAM in the generic kernel and now can't understand NCQ works or not. options ATA_CAM enables wrapper for existing ata(4) controller drivers, that do not support command queuing. All this option give to you is switching to newer bus management and peripheral code. ada0 at ata2 bus 0 scbus2 target 0 lun 0 ada0: ST31000524NS SN11 ATA-8 SATA 2.x device ada0: 300.000MB/s transfers (SATA 2.x, UDMA5, PIO 8192bytes) ada0: 953869MB (1953525168 512 byte sectors: 16H 63S/T 16383C) If NCQ was supported, you would also see there lines: ada0: Command Queueing enabled This system installed on Tyan Tomcat K8E motherboard. To support MCQ you should use one of new drivers: ahci(4), siis(4) or mvs(4) on supported hardware. As I can see, your board is based on nForce4 chipset, which supports NCQ in own proprietary way, which is not supported at the moment and unlikely will be ever supported, as it is quite old already and all later chisets are AHCI compatible instead of it. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: 8.0-RELEASE/amd64 - full ZFS install - low read and write disk performance
Artem Belevich wrote: aoc-sat2-mv8 was somewhat slower compared to ICH9 or LSI1068 controllers when I tried it with 6 and 8 disks. I think the problem is that MV8 only does 32K per transfer and that does seem to matter when you have 8 drives hooked up to it. I don't have hard numbers, but peak throughput of MV8 with 8-disk raidz2 was noticeably lower than that of LSI1068 in the same configuration. Both LSI1068 and MV2 were on the same PCI-X bus. It could be a driver limitation. The driver for Marvel SATA controllers in NetBSD seems a bit more advanced compared to what's in FreeBSD. I also wouldn't recommend to use Marvell 88SXx0xx controllers now. While potentially they are interesting, lack of documentation and numerous hardware bugs make existing FreeBSD driver very limited there. I wish intel would make cheap multi-port PCIe SATA card based on their AHCI controllers. Indeed. Intel on-board AHCI SATA controllers are fastest from all I have tested. Unluckily, they are not producing discrete versions. :( Now, if discrete solution is really needed, I would still recommend SiI3124, but with proper PCI-X 64bit/133MHz bus or built-in PCIe x8 bridge. They are fast and have good new siis driver. On Mon, Jan 25, 2010 at 3:29 AM, Pete French petefre...@ticketswitch.com wrote: I like to use pci-x with aoc-sat2-mv8 cards or pci-e cardsthat way you get a lot more bandwidth.. I would goalong with that - I have precisely the same controller, with a pair of eSATA drives, running ZFS mirrored. But I get a nice 100 meg/second out of them if I try. My controller is, however on PCI-X, not PCI. It's a shame PCI-X appears to have gone the way of the dinosaur :-( -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: 8.0-RELEASE/amd64 - full ZFS install - low read and write disk performance
Dan Naumov wrote: Alexander, since you seem to be experienced in the area, what do you think of these 2 for use in a FreeBSD8 ZFS NAS: http://www.supermicro.com/products/motherboard/ATOM/ICH9/X7SPA.cfm?typ=H http://www.supermicro.com/products/motherboard/ATOM/ICH9/X7SPA.cfm?typ=HIPMI=Y Unluckily I haven't yet touched Atom family close yet, so I can't say about it's performance. But higher desktop level (even bit old) ICH9R chipset there is IMHO a good option. It is MUCH better then ICH7, often used with previous Atoms. If I had nice small Mini-ITX case with 6 drive bays, I would definitely look for some board like that to build home storage. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: 8.0-RELEASE/amd64 - full ZFS install - low read and write disk performance
Chris Whitehouse wrote: Dan Naumov wrote: CPU-performance-wise, I am not really worried. The current system is an Atom 330 and even that is a bit overkill for what I do with it and from what I am seeing, the new Atom D510 used on those boards is a tiny bit faster. What I want and care about for this system are reliability, stability, low power use, quietness and fast disk read/write speeds. I've been hearing some praise of ICH9R and 6 native SATA ports should be enough for my needs. AFAIK, the Intel 82574L network cards included on those are also very well supported? These might be interesting then www.fit-pc.com The Intel US15W SCH chipset or System Controller Hub as it's called is mentioned in hardware notes for 8.0R and 7.2R but only for snd_hda, I don't know if this means other functions are supported or not. This thread says it is supported http://mail-index.netbsd.org/port-i386/2010/01/03/msg001695.html Intel US15W (SCH) chipset heavily stripped and tuned for netbooks. It has no SATA, only one PATA channel. It is mostly supported by FreeBSD, but with exception of video, which makes it close to useless. it has only one benefit - low power consumption. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: PCIe audio cards: what is tob be preferred with FreeBSD 8.0/9-CURRENT?
O. Hartmann wrote: Well, At this very moment I utilise a M-Audio 5.1 PCI-audio board with which I'm really satisfied. My next box doesn't have PCI slots at all (ASUS P6T6-WS Revolution) and due to the fact I'm using Windows 7 sometimes for recreational gaming, I'd like to have a moderate expensive audio board with the workstation which is supported by FreeBSD 8/9. In the past - means two or three ywars ago, I had problems with Soundblaster PCIe boards, so I was recommended avoiding those and choosing the more elabotrated M-Audio cards for the PCI bus. At this moment, I look for the Soundblaster X-Fi range of PCIe cards, but I'm not sure whether they are supported by FreeBSd 8/9. Any suggestions? I would first test on-board HDA codec, placed on that motherboard. They are free and usually work with FreeBSD just out of the box, using snd_hda driver. Now snd_hda and snd_uaudio are only two drivers supporting multichannel playback. If you need analog connection, snd_hda can usually provide multichannel 24bit/192kHz playback. But also it supports digital SPDIF I/O, including AC3/DTS pass-through. Together with SPDIF-connected external audio receiver, even simple HDA codec could become very interesting high-quality choice. Personally I am completely fulfilled with combination of simple Realtek HDA codec, digitally connected via SPDIF to Marantz SR4001 receiver, loaded to the full-sized Eltax 7.1 speaker set. Previously I was using Creative Audigy2 ZS, but now it just collecting dust in my table. It works, but I really don't need it. PS: If you want to look cool, you may use optical SPDIF connection. :) -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: 8.0-RELEASE/amd64 - full ZFS install - low read and write disk performance
Dan Naumov wrote: This works out to 1GB in 36,2 seconds / 28,2mb/s in the first test and 4GB in 143.8 seconds / 28,4mb/s and somewhat consistent with the bonnie results. It also sadly seems to confirm the very slow speed :( The disks are attached to a 4-port Sil3124 controller and again, my Windows benchmarks showing 65mb/s+ were done on exact same machine, with same disks attached to the same controller. Only difference was that in Windows the disks weren't in a mirror configuration but were tested individually. I do understand that a mirror setup offers roughly the same write speed as individual disk, while the read speed usually varies from equal to individual disk speed to nearly the throughput of both disks combined depending on the implementation, but there is no obvious reason I am seeing why my setup offers both read and write speeds roughly 1/3 to 1/2 of what the individual disks are capable of. Dmesg shows: atapci0: SiI 3124 SATA300 controller port 0x1000-0x100f mem 0x90108000-0x9010807f,0x9010-0x90107fff irq 21 at device 0.0 on pci4 ad8: 1907729MB WDC WD20EADS-32R6B0 01.00A01 at ata4-master SATA300 ad10: 1907729MB WDC WD20EADS-00R6B0 01.00A01 at ata5-master SATA300 8.0-RELEASE, and especially 8-STABLE provide alternative, much more functional driver for this controller, named siis(4). If your SiI3124 card installed into proper bus (PCI-X or PCIe x4/x8), it can be really fast (up to 1GB/s was measured). -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: 8.0-RELEASE/amd64 - full ZFS install - low read and write disk performance
Dan Naumov wrote: On Mon, Jan 25, 2010 at 2:14 AM, Dan Naumov dan.nau...@gmail.com wrote: On Sun, Jan 24, 2010 at 11:53 PM, Alexander Motin m...@freebsd.org wrote: Dan Naumov wrote: This works out to 1GB in 36,2 seconds / 28,2mb/s in the first test and 4GB in 143.8 seconds / 28,4mb/s and somewhat consistent with the bonnie results. It also sadly seems to confirm the very slow speed :( The disks are attached to a 4-port Sil3124 controller and again, my Windows benchmarks showing 65mb/s+ were done on exact same machine, with same disks attached to the same controller. Only difference was that in Windows the disks weren't in a mirror configuration but were tested individually. I do understand that a mirror setup offers roughly the same write speed as individual disk, while the read speed usually varies from equal to individual disk speed to nearly the throughput of both disks combined depending on the implementation, but there is no obvious reason I am seeing why my setup offers both read and write speeds roughly 1/3 to 1/2 of what the individual disks are capable of. Dmesg shows: atapci0: SiI 3124 SATA300 controller port 0x1000-0x100f mem 0x90108000-0x9010807f,0x9010-0x90107fff irq 21 at device 0.0 on pci4 ad8: 1907729MB WDC WD20EADS-32R6B0 01.00A01 at ata4-master SATA300 ad10: 1907729MB WDC WD20EADS-00R6B0 01.00A01 at ata5-master SATA300 8.0-RELEASE, and especially 8-STABLE provide alternative, much more functional driver for this controller, named siis(4). If your SiI3124 card installed into proper bus (PCI-X or PCIe x4/x8), it can be really fast (up to 1GB/s was measured). -- Alexander Motin Sadly, it seems that utilizing the new siis driver doesn't do much good: Before utilizing siis: iozone -s 4096M -r 512 -i0 -i1 random randombkwd record stride KB reclen write rewritereadrereadread writeread rewrite read fwrite frewrite fread freread 4194304 512 28796 287665161050695 After enabling siis in loader.conf (and ensuring the disks show up as ada): iozone -s 4096M -r 512 -i0 -i1 random randombkwd record stride KB reclen write rewritereadrereadread writeread rewrite read fwrite frewrite fread freread 4194304 512 28781 288974721450540 Just to add to the numbers above, exact same benchmark, on 1 disk (detached 2nd disk from the mirror) while using the siis driver: random randombkwd record stride KB reclen write rewritereadrereadread writeread rewrite read fwrite frewrite fread freread 4194304 512 57760 563716886774047 If both parts of mirror uses same controller, it doubles it's bus traffic. That may reduce bandwidth twice. The main benefit of siis(4) is a command queuing. You should receive more benefits on multithread random I/O. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: FreeBSD 8.0 STABLE sata regression
Hi. Abdullah Ibn Hamad Al-Marri wrote: My servers are alright with FreeBSD 8.0-RELEASE, but not with -STABLE. FreeBSD NOC.WEArab.Net 8.0-STABLE FreeBSD 8.0-STABLE #0: Fri Jan 8 18:26:13 UTC 2010 ad...@localhost.localdomain:/usr/obj/usr/src/sys/GENERIC amd64 atapci0: Intel ICH7 SATA300 controller port 0x1f0-0x1f7,0x3f6,0x170-0x177,0x376,0x20a0-0x20af mem 0x9020-0x902003ff irq 19 at device 31.2 on pci0 ata0: ATA channel 0 on atapci0 ata0: [ITHREAD] ata1: ATA channel 1 on atapci0 ata1: [ITHREAD] when I upgraded to FreeBSD 8.0-STABLE it got many issues. ad2: 381554MB WDC WD4000KD-00NAB0 01.06A01 at ata1-master UDMA100 If UDMA mode reported instead of SATA speed is the only issue, then it is not an issue, but only a cosmetic change. It doesn't mean that your drives will work slower. Thanks for report, I'll plug ICH7 board a bit later to check this. To get more info about the problem, try `atacontrol mode ad2` command. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: FreeBSD 8.0 STABLE sata regression
Abdullah Ibn Hamad Al-Marri wrote: Abdullah Ibn Hamad Al-Marri wrote: My servers are alright with FreeBSD 8.0-RELEASE, but not with -STABLE. FreeBSD NOC.WEArab.Net 8.0-STABLE FreeBSD 8.0-STABLE #0: Fri Jan 8 18:26:13 UTC 2010ad...@localhost.localdomain:/usr/obj/usr/src/sys/GENERIC amd64 atapci0: port 0x1f0-0x1f7,0x3f6,0x170-0x177,0x376,0x20a0-0x20af mem 0x9020-0x902003ff irq 19 at device 31.2 on pci0 ata0: on atapci0 ata0: [ITHREAD] ata1: on atapci0 ata1: [ITHREAD] when I upgraded to FreeBSD 8.0-STABLE it got many issues. ad2: 381554MB at ata1-master UDMA100 If UDMA mode reported instead of SATA speed is the only issue, then it is not an issue, but only a cosmetic change. It doesn't mean that your drives will work slower. Thanks for report, I'll plug ICH7 board a bit later to check this. To get more info about the problem, try `atacontrol mode ad2` command. Here is the result, VB# atacontrol mode ad4 current mode = UDMA100 VB# atacontrol mode ad6 current mode = UDMA100 VB# OK. I have attached my ICH7 board and found where is the problem. This chipset doesn't provide access to SATA control registers. As result, driver can't get SATA connection info and so doesn't report it. If you wish, you can restore previous behavior by applying attached patch. Here is what I have with it: %atacontrol mode ad0 current mode = UDMA100 %atacontrol mode ad4 current mode = UDMA100 SATA 1.5Gb/s %atacontrol mode ad5 current mode = UDMA100 SATA 1.5Gb/s %atacontrol mode ad6 current mode = UDMA100 SATA 1.5Gb/s %atacontrol mode ad7 current mode = UDMA100 SATA 1.5Gb/s But as I have said before, it is only a cosmetics. -- Alexander Motin --- ata-pci.c.prev 2010-01-10 01:25:45.0 +0200 +++ ata-pci.c 2010-01-10 01:25:49.0 +0200 @@ -714,9 +714,12 @@ static int ata_pcichannel_getrev(device_t dev, int target) { struct ata_pci_controller *ctlr = device_get_softc(device_get_parent(dev)); + struct ata_channel *ch = device_get_softc(dev); if (ctlr-getrev) return (ctlr-getrev(dev, target)); + else if (ch-flags ATA_SATA) + return (1); else return (0); } ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: Phoronix Benchmarks: Waht's wrong with FreeBSD 8.0?
O. Hartmann wrote: I'm just wondering what's wrong with FreeBSD 8.0/amd64 when I read the Benchmarks on Phoronix.org's website. Especially FreeBSD's threaded I/O shows in contrast to all claims that have been to be improoved the opposite. Instead of trying to compare something, I propose to look on that numbers itself first: - first test tells that average write latency is about 100us. But it looks quite surprising for Laptop HDD, which has seek time of at least several milliseconds. - second test - a bit closer to life - 2-3ms - ok, Linux won here slightly, as FreeBSD installation in this test had no NCQ support. - third test - 9us per write on Linux. I am just crying. - forth test - all OSes gave 50-80us. Probably it is just a buffer case read time. So most of shown cases are testing almost only file system cache parameters. It is just insane to compare them for so different systems with so different write-back policies. If somebody still have questions, after some UFS parameters tuning I've got with the same tiotest tool: - Random Write latency - 15us, - Random Read latency - 7us. So who can beat my FreeBSD? :))) What's about second test. To check possible NCQ effect I've built test setup with new 320GB 7200RPM Seagate drive connected to Intel ICH10R controller. I've run IMHO more reasonable benchmark/raidtest tool from ports on whole device, to execute pregenerated random mix of 1 random-sized (512B - 128KB) read/write requests using default ata(4) driver and new ahci(4): Number of READ requests: 5029. Number of WRITE requests: 4971. Number of bytes to transmit: 655986688. Number of processes: 32. The results: ata(4) - no NCQ: Bytes per second: 12455402 Requests per second: 189 ahci(4) - with NCQ: Bytes per second: 19889778 Requests per second: 303 Results are repeatable up to the 4-th digit. Average time per request is 5.29ms and 3.3ms respectively, that is realistic for this drive. So, with such difference, I believe, we will not loose this test any more. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: No sound from snd_hda
Steve Franks wrote: Should I be asking you about the old snd_hda? I'm having this same problem and ignoring it for some time (no sound), but on FreeBSD aire.franks-development.dyndns.biz 6.4-STABLE FreeBSD 6.4-STABLE #9: Sat Dec 13 18:25:36 MST 2008 r...@aire.franks-development.dyndns.biz:/usr/obj/usr/src/sys/GENERIC amd64 instead of current. Just take HDA driver from CURRENT, it works fine on RELENG_7 and I think it may work on your RELENG_6. I read the warning: A few Hardware/OEM vendors tend to screw up BIOS settings, thus rendering the snd_hda driver useless, which usually results in a state where the snd_hda driver seems to attach and work, but without any sound. Such unresolved cases are happen sometimes, but they are really rare. I have being reported only 2 or 3 of them while working on new driver. Absolute majority of devices working fine out of the box, or sometimes require minimal tuning to fix BIOS issues. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to freebsd-questions-unsubscr...@freebsd.org
Re: Can't get soundcard to work
Frank Shute wrote: On Thu, Oct 09, 2008 at 09:12:28AM +0200, Aniruddha wrote: On Wed, 2008-10-08 at 23:31 +0100, Frank Shute wrote: ls /dev/d da0 da2 dconsdevstat dsp0.0 dsp1.0 da1 da3 devctl dgdb dsp0.1 dumpdev On my machine: $ ls /dev | grep dsp dsp0.0 dsp0.1 i.e no dsp1.0 It will be created on demand when sound application will try to open it directly. Also it may be set as default with hw.snd.default_unit sysctl described in man sound. What does: $ cat /dev/sndstat give you? sndstat: FreeBSD Audio Driver (newpcm: 32bit 2007061600/i386) Installed devices: pcm0: ATI (Unknown) High Definition Audio Controller at memory 0xff9ec000 irq 17 kld snd_hda [20080420_0052] [MPSAFE] (mixer only) pcm1: Intel 82801G High Definition Audio Controller at memory 0xffafc000 irq 19 kld snd_hda [20080420_0052] [MPSAFE] (1p:1v/1r:1v channels duplex) The driver is binding to both. Indeed. First one is HDMI on video and second on motherboard. The old snd_hda driver you have surely unable to manage first one. Second looks better, but also should be tested. Updated driver in 8-CURRENT should work better. It could be ported to 7-STABLE without any modifications. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: clone a drive, no raid involved
Brad Mettee wrote: I'm setting up a pair of machines with almost identical OS config, and completely identical hardware. One is a primary DNS server, the other is secondary. NS1 will also serve web, NS2 will be a mail server. Both are low volume/loads. It looks like I can use DD to copy an entire drive, but it's a 500G drive and that's going to take a really long time (especially since it's brand new with no data besides base OS). My question: Is there a better way to duplicate a drive including boot info? Partition second drive (I usually prefer sysinstall which also installs loaders), mount it and use dump/restore: cd /mnt dump -aLf - -C32 / | restore -rf - -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: FreeBSD+mpd freeze after connection establishing
Alexey A. Ukhov wrote: I have problem with subj. After start mpd as client and connecting to other mpd-server FreeBSD freeze. No any log messages, kernel panics, etc. One of the most popular PPTP client problems is the routing loop created by using same server address inside and outside of tunnel. In such case address inside of tunnel with /32 netmask usually more preferred for the system and it makes all tunnel traffic wrap inside tunnel itself causing unpredictable results. Addresses inside and outside of tunnel _MUST_BE_DIFFERENT_! In some cases problem can be workarounded with some firewall forwarding or some other alike techniques. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: mpd5 stoped working after subnet change
Reinhold wrote: So, upon connection, there was a bandwidth of 64000bps, but upon closing that session, the b/w is 9600bps. I see the same things since I started logging the connection. At the start I see this mpd5: [B-3] Bundle: Status update: up 1 link, total bandwidth 64000 bps and at the end its mpd5: [B-3] Bundle: Status update: up 0 links, total bandwidth 9600 bps Those speeds mean nothing to you actually. They are used _only_ for multilink load balancing. 64000 is the default that can be changed manually by configuration, but as I have said it is meaningless whan multilink is not used. PPTP protocol has capabilities to make PAC inform PNS side about connection speed, but it is only informational and in no way interfere with real transfer speeds. This is a bit OT because my original problem got fixed thanks to you, I still have no idea what fixed it though, hehehe Logging (probably increased to 'log +phys2 +phys3') and tcpdump on a physical interface should help you if problem will get back again. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: SATA support custom v7.0 kernel
Colin Brace пишет: I'd like to add a SATA drive to a P4 box that I use as a gateway/server. The MB doesn't have a SATA interface, so I stuck an Initio-based SATA card in the box. The kernel recognizes the card, but the attached drive wasn't displayed when I started up sysinstall to format it. I'm wondering whether I am missing SATA support. I commented out a bunch of stuff such as the SCSI controllers when I compiled the kernel with ALTQ support, but these options *are* included: # SCSI peripherals device scbus # SCSI bus (required for SCSI) device ch # SCSI media changers device da # Direct Access (disks) device sa # Sequential Access (tape etc) device cd # CD device pass # Passthrough device (direct SCSI access) device ses# SCSI Environmental Services (and SAF-TE) Should this be enough? I don't see any specific references to SATA in the handbook http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/kernelconfig-config.html Usually SATA is more ATA then SCSI. Try to add to your config: device ata device atadisk # ATA disk drives device ataraid # ATA RAID drives device atapicd # ATAPI CDROM drives device atapifd # ATAPI floppy drives device atapist # ATAPI tape drives -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Crypto on motherboard
Harry Matthiesen Jensen wrote: After I loaded a driver (ubsec), I saw this line in the log: cryptosoft0: software crypto on motherboard Is it something I can take advantage of? This is software cryptography pseudo-driver working without any real hardware acceleration. It is used by IPSec, geli and some other subsystems when there is no any hardware cryptography accelerator present in system. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: mpd pptp server?
On Wed, 19 Mar 2008 00:43:58 +0100 Jon Theil Nielsen [EMAIL PROTECTED] wrote: Do I need to have a customized kernel to make it work? Or are there any obvious errors in the above configuration? Mpd4 should work without special system tuning. The best way to find the problem is to read it's logs. Mpd writes detailed logs using syslog (you should configure syslog.conf for it alike to ppp) and to the stdout if running in foreground. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: Inet access via serial interface
Potocki, Mariusz wrote: ps. Two radiomodems are invisible and act as a vry lng null-modem cable. Windows treats NULL-modem connection as special case of dial-up, when each side has some specific chat script. AFAIR it is something like 'CLIENT CLIENTSERVER'. Google should help you. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: mpd with a dual pppoe setup
Reinhold wrote: If you are connecting to the ISP it would be better not to specify IP to let them be negotiated. How will I go ahead and set the static ips for both my wan connections? If IPs are static then provider itself will negotiate them every time same. If you will specify them and provider will request another, negotiation will fail. So I would not recommend you to specify them. For wan1 I have one ip and for wan2 I have 5 ips Probably for wan2 you have one interface IP plus some network routed to you. PPP protocol does not allows to negotiate routed network, so that network probably should be configured by hands somewhere. I was orginally thinking of seting up proxy-arp to add the extra 4 ips on wan2 after connecting using the first ip, but now I'm not to sure anymore. proxy-arp is working only with Ethernet. If you are going to give that static IPs only to PPTP users - you probably don't need it. We need these ips because some of the websites and servers that we manage are ip restricked for added security, this is also the reason why I have a pptp server for the people thats not in our office. You can specify static IPs for those users in mpd.secret file. How does mpd5 assign the ng interfaces, in my previose cong there was a line new -i ng0 provider PPPoE0, does mpd5 do this automatically? Interfaces will be allocated by system in order of usage. WAN1 will be ng0, WAN2 - ng1 and the PPTP link's interfaces will be allocated dynamically on connect. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: mpd with a dual pppoe setup
Hi. Reinhold wrote: Here is what I've done so far. /usr/local/etc/mpd5/mpd.conf The config you have made probably suits to mpd3/4, but not to mpd5. Mpd5 uses alike commands, but general configuration idea is different. There is no mpd.links file in mpd5. Read new examples present in package. default: load wan1 load wan2 Then it would be reasonable to merge pptpd: section here. set ipcp ranges static-ip-0/32 isp-gateway-0/32 If you are connecting to the ISP it would be better not to specify IP to let them be negotiated. load common_setting wan2: new -i ng1 wan2 PPPoE1 set iface route default set iface up-script /usr/local/etc/mpd/script-wan2.sh set iface down-script /usr/local/etc/mpd/script-wan2.sh set bundle authname [EMAIL PROTECTED] set bundle password passwd set ipcp ranges static-ip-1/32 isp-gateway-1/32 load common_setting common_setting: set iface addrs 1.1.1.1 2.2.2.2 set iface disable on-demand Without on-demand set iface addrs is useless. set iface idle 0 This is default. # PPTP pt0: set link type pptp set pptp enable incoming set pptp disable originate set pptp disable windowing set pptp self 127.0.0.1 It will be difficult to accept incoming while listening on 127.0.0.1. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: mpd4 cann't work after upgrade freebsd os
Pei Pjf wrote: mail# kldload ng_socket kldload: can't load ng_socket: No such file or directory mail# kldload ng_iface kldload: can't load ng_iface: No such file or directory mail# kldload ng_tty kldload: can't load ng_tty: No such file or directory mail# kldload ng_ppp kldload: can't load ng_ppp: No such file or directory mail# kldload ng_iface kldload: can't load ng_iface: No such file or directory mail# kldload ng_vjc kldload: can't load ng_vjc: No such file or directory mail# I think these files were lost when upgrading os. I will search these from another freebsd box. They may be not lost but unreachable due to incorrect module search path. Try something like `kldload /boot/kernel/netgraph.ko`. If it will work, then you may be needed to update your /boot/loader.rc from the /usr/src/sys/boot/i386/loader/loader.rc. I had alike problem with modules loading on several boxes after several source upgrades and I have fixed it in a such way. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
MPPC compression implementations legal status?
Hi. Can anybody explain me for sure current license status of Microsoft Point-to-Point Compression (MPPC) Protocol? It is not implemented in FreeBSD for years and I think it would be good to change this. In RFC 2118 told that Source and object licenses are available on a non-discriminatory basis from Stac Electronics. Does it means requirement to get license to use HIFN implementation of this protocol or also denies any other free implementation? How does it coexist with Microsoft's ROYALTY FREE PROTOCOL LICENSE AGREEMENT, http://msdn2.microsoft.com/en-us/library/ms818807.aspx. Doesn't this agreement allows free implementation of server side protocols for cooperation with Windows client systems? -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]
Re: mpd 4.0b5 does not want to work
Hi. [EMAIL PROTECTED] wrote: I have compiled and installed mpd 4.0b5 on my PC with FreeBSD 6.1-RELEASE. Then I created the next files: # mpd.conf open iface Due to b5 internal rework, syntax open iface is now deprecated in it's original meaning. Now it can be used only for Dial-On-Demand implemantation. You should use open lcp or better just open command. -- Alexander Motin ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to [EMAIL PROTECTED]