From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: the one debian. On Wednesday, July 23, 2003, at 10:10 AM, Craig Dooley wrote: -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I definitely agree that the user system should be built using packages. People dont care about downloading a 300meg source tree and waiting an hour to get a new release. Yes rpm systems are a hassle to upgrade, but debian has shown that you dont need source for a system. Debian allows someone to do dpkg --get-selections > profile and have a list of programs to duplicate a current working system. I like the environment idea to allow multiple instances of needed dependancies (it would fix things like debian with libpng or freebsd with freetype) but unless I am missing something, we're still gonna need a build system to create these packages. Also, a database backed approach seems the easies way to conserve bandwidth and allow full updates. In my opinion, in the end, we should have a "buildworld" just rebuild all the base packages and update them that way. People who care about the source can get it just as easily, and people that dont won't ever have to worry. Also, will ports stay with just a new backend for building packages? Will there be a new build environment? A binary downloading frontend seems a necessity now also. Just my thoughts On Tuesday 22 July 2003 05:32 pm, Thierry Herbelot wrote: Matthew Dillon wrote: The eventual goal is to make *ALL* userland applications, including things like cp, ls, etc... all operate through the packaging system, and remember that it is the packaging system's goal to support installation of multiple versions of anything without conflict. oh no ! please ! a fine-grained packaging system used for the core OS is just a pain : I'm fed up with DeathTrap or Mandrake or Suse or .... with their thousands of rpm's - how do you upgrade such a machine ? (with something intelligent like portupgrade ? but there are lots of oldish shared libs left behind) - how do you duplicate a machine ? (preferably with some automatic procedure) after living some time, it's very difficult to know the smallest set of base rpms you have to select to get the same install (with the other packages going in via dependencies) One big selling point of FreeBSD is the relative ease of configuration description : get the full OS, then add a smallish number of outside packages. TfH - -- - -Craig cd5697 at xxxxxxxxxx -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.2 (FreeBSD) iD8DBQE/HdJhYXswu1EOVs0RAj4iAJsHi+qRvm90VPGUpTzNbz0QZydCegCglkE3 3OgUeRMZaa5sq/OWVtMr7lY= =lGnj -----END PGP SIGNATURE----- From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: this in about the same way that OpenAFS does volume-mounting, which is what I was wondering. Thanks. -- Garance Alistair Drosehn = gad at xxxxxxxxxxxxxxxxxxxx Senior Systems Programmer or gad at xxxxxxxxxxx Rensselaer Polytechnic Institute or drosih at xxxxxxx From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: P4 3.06ghz from buildworld timing. Also most of the numbers are a lot lower on the AMD side as far as memory operations and context switches which seems like a big benefit. (2) The AMD's syscall overhead is very low compared to the P4. I ran the 'sc1' test in /usr/src/test/sysperf. The AMD box comes back at 242nS/loop while the P4 takes 1300nS/loop and the DELL box (LEAF/SMP) takes 617nS/loop. getuid() 0.916s 2402600 loops = 0.381uS/loop on my AMD 1.4ghz. I think big reason for the speed improvement here is because on every syscall entry on a P4 you hit a branch misprediction and the AMD it cares more about just blowing through the instructions which makes it a lot faster at most things. -DR From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: seems to be running fine on FreeBSD-STABLE. Does that still apply to DragonFly? I believe there's some kernel module cruft in there, so my bet is that it _might_ work after some kludging in the module sources. I'm downloading an OpenAFS snapshot as we speak (their webserver isn't that fast. sigh), and I'll try to get it going. I'll report success and I'll submit patches if need be. (Disclaimer: I'm no AFS expert. I'm only doing this because I currently have the need for an AFS server, and don't feel like going with Linux) On a side (but not really unrelated) note, since we don't have KerberosIV anymore, might it be an idea to rename all the kerberos5 stuff back to their respective names instead of that k5* mess? Also, it would be really nice if the kerberos keytabs and configs would reside in /etc/kerberosV instead of plain /etc. Cheers, Emiel -- From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: quite similar to what we have implemented in K42, and naturally this makes us quite excited about the directions the dragonflybsd effort intends to take. We'd be very interested in having ongoing interactions with this community to share ideas, experiences, horror stories and code. Consequently, I would like to invite members of this community to check out the K42 web-site http://www.research.ibm.com/K42 where we have numerous papers that describe our work. Some of the papers that I think are most relevant: An overview paper: http://www.research.ibm.com/K42/white-papers/Overview.pdf A paper describing how K42 provides traditional UNIX API's: http://www.research.ibm.com/K42/papers/freenix03.pdf A paper describing K42's threading and scheduling infrastructure: http://www.research.ibm.com/K42/white-papers/Scheduling.pdf Please bear in mind that K42 development is ongoing and consequently static papers may be out of date. Nonetheless, these papers should be an adequate starting point to understanding K42 and hopefully a discussion of common interests and approaches. I will be happy to answer any questions you may have. Regards, Michal Ostrowski mostrows at xxxxxxxxxxxxxx From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: 2003 Disabling a Local APIC Disables Both Logical Processor APICs on a Hyper-Threading Technology Enabled Processor Problem: Disabling a local APIC on one logical processor of a Hyper-Threading Technology enabled processor by clearing bit 11 of the IA32_APIC_BASE MSR will effectively disable the local APIC on the other logical processor Implication: Disabling a local APIC on one logical processor prevents the other logical processor from sending or receiving interrupts. Multiprocessor Specification compliant BIOSs and multiprocessor operating systems typically leave all local APICs enabled preventing any end-user visible impact from this erratum. Workaround: Do not disable the local APICs in a Hyper-Threading Technology enabled processor. . .. Problem: When a system bus agent (processor or chipset) issues an interrupt transaction without data onto the system bus and the transaction receives a HardFailure response, a potential processor hang can occur. . .. From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: Pentium Pro MTRR support enabled md0: Malloc disk Using $PIR table, 12 entries at 0xc00f4660 acpi0: on motherboard installed MI handler for int 9 acpi0: power button is handled as a fixed feature programming model. Timecounter "ACPI-fast" frequency 3579545 Hz acpi_timer0: <24-bit timer at 3.579545MHz> port 0x408-0x40b on acpi0 acpi_cpu0: on acpi0 acpi_cpu1: on acpi0 acpi_button0: on acpi0 npx0: on motherboard npx0: INT 16 interface pcib0: on motherboard We stop booting here. Here is the output from pciconf -lv none0 at pci0:0:0: class=0x060000 card=0x25608086 chip=0x25608086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82845G/GL/GV/GE/PE DRAM Controller / Host-Hub I/F Bridge' class = bridge subclass = HOST-PCI agp0 at pci0:2:0: class=0x030000 card=0x4c598086 chip=0x25628086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82845G/GL/GV/GE/PE Integrated Graphics Device' class = display subclass = VGA uhci0 at pci0:29:0: class=0x0c0300 card=0x4c598086 chip=0x24c28086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801DB (ICH4) USB UHCI Controller #1' class = serial bus subclass = USB uhci1 at pci0:29:1: class=0x0c0300 card=0x4c598086 chip=0x24c48086 rev=0x01 hdr=0x00 guppy% cat stuff none0 at pci0:0:0: class=0x060000 card=0x25608086 chip=0x25608086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82845G/GL/GV/GE/PE DRAM Controller / Host-Hub I/F Bridge' class = bridge subclass = HOST-PCI agp0 at pci0:2:0: class=0x030000 card=0x4c598086 chip=0x25628086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82845G/GL/GV/GE/PE Integrated Graphics Device' class = display subclass = VGA uhci0 at pci0:29:0: class=0x0c0300 card=0x4c598086 chip=0x24c28086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801DB (ICH4) USB UHCI Controller #1' class = serial bus subclass = USB uhci1 at pci0:29:1: class=0x0c0300 card=0x4c598086 chip=0x24c48086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801DB (ICH4) USB UHCI Controller #2' class = serial bus subclass = USB uhci2 at pci0:29:2: class=0x0c0300 card=0x4c598086 chip=0x24c78086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801DB (ICH4) USB UHCI Controller #3' class = serial bus subclass = USB none1 at pci0:29:7: class=0x0c0320 card=0x4c598086 chip=0x24cd8086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801DB (ICH4) USB EHCI Controller' class = serial bus subclass = USB pcib1 at pci0:30:0: class=0x060400 card=0x00000000 chip=0x244e8086 rev=0x81 hdr=0x01 vendor = 'Intel Corporation' device = '82801BA/CA/DB (ICH2/3/4) Hub Interface to PCI Bridge (244E)' class = bridge subclass = PCI-PCI isab0 at pci0:31:0: class=0x060100 card=0x00000000 chip=0x24c08086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801DB (ICH4) LPC Interface Bridge' class = bridge subclass = PCI-ISA atapci0 at pci0:31:1: class=0x01018a card=0x4c598086 chip=0x24cb8086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801DB (ICH4) UltraATA/100 EIDE Controller' class = mass storage subclass = ATA none2 at pci0:31:3: class=0x0c0500 card=0x4c598086 chip=0x24c38086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801DB (ICH4) SMBus Controller' class = serial bus subclass = SMBus none3 at pci0:31:5: class=0x040100 card=0x03028086 chip=0x24c58086 rev=0x01 hdr=0x00 vendor = 'Intel Corporation' device = '82801DB (ICH4) AC'97 Audio Controller' class = multimedia subclass = audio fxp0 at pci1:8:0: class=0x020000 card=0x30128086 chip=0x10398086 rev=0x81 hdr=0x00 vendor = 'Intel Corporation' device = '82801DB (ICH4) PRO/100 VE Network Connection' class = network subclass = ethernet -DR From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: We are pleased to announce that we have chosen a new file system to use in SkyOS, and it will be available in SkyOS 5.0. The file system that we have chosen to use is the OpenBFS, made available under the MIT license by the OpenBeOS project. We have chosen this excellent It's interesting, though I'd say, having used BeOS and therefore BFS for a few years, you'd want to pick this filesystem clone because of the attribute support, not because of journaling. File attributes was really neat, and gave the filesystem database-like capabilities, but it's not something that will mesh easily with 99% of the unixy tools out there, so it almost could be considered "baggage". From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: about anything you'd want to use as a UI. From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: XML doesn't care about the presentation of the data, just the data itself. It would be up to the UI to worry about displaying the data, asking for user input, that type of thing. So, following an MS analogy, the 'wizard' logic would be in the UI, get/set-edit/create/command would be passed as XML formatted messages, and you'd get XML formatted asynchronous responses and status reports arriving back from the agent. Keeping the interface asynchronous buys you a lot of flexibility in UI-wise, I think. Andrew. From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: DragonFly's weak spot. I'm guessing that this is not a problem with DragonFly, per se, as much as it is simply that the group has put most of its effort into other areas of the OS and hasn't had the resources to address the issue. From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: know WINE cannot run well on FreeBSD 4/5 due to VM issue. So what about DragonFly? http://www.winehq.org/?issue=230#Wine%20on%20FreeBSD ----------------------------------------------------------------- 06/09/2004 Wine on FreeBSD Archive Ports Remember last week how I alluded to a problem with Wine and FreeBSD? That thread began about a month ago and it started when John Birrell wondered how well Wine worked on FreeBSD-CURRENT (5.2.1) compared to -STABLE (4.10). Jonathan Fosburgh described the current (pun intended) situation, There is a threading problem right now (search the bug reports) on -CURRENT when using libpthread. Using libc_r is sort of working, at least as of a few weeks ago. I saw some discussion on one of the FreeBSD mailing lists discussing a different way of linking against a thread library that might make things work better. Basically, it involves not using -lpthread during the link stage but instead compiling against -pthread. Check through the freebsd-current and freebsd-ports mailing list archives, I believe the discussion was in one of those. Gerald Pfeifer, the FreeBSD packager, then brought up the fact that -STABLE doesn't work either: Right now, Wine doesn't work at all on FreeBSD -STABLE: wine: failed to initialize: /swtest/wine/dlls/ntdll.dll.so: mmap of entire address space failed: Cannot allocate memory and before that I used to see deadlocks upon startup of non-trivial applications (such as Forte Agent, both 16bit and 32bit flavors). I believe there are also significant threading issues on -CURRENT, so overall Wine is hardly, if at all, usable on any version of FreeBSD I have access to, even though I'm still working to keep it at least compilable on FreeBSD 4.9 and 5.2/5.3. John then looked into the problem and reported what he found: From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: CVS on FreeBSD -current that seem to be related to the use of BSD make. In the dlls 'setupapi' and 'version' directories there are source files called install.c. These confuse the build during a "make install". BSD make tries to compile and link a program called "install". Renaming the source files to something other than install.c (such as winstall.c) allows the build to succeed. At run time, the error: "wine: failed to initialize: /something/lib/wine/ntdll.dll.so: mmap of entire address space failed: Cannot allocate memory" is caused by Wine attempting to mmap memory outside the user process address space. I see mmap addr around 0xd8100000 (mostly 0xd81eccd8) whereas the user address space limit on FreeBSD current is 0xbfc00000 (at least on my system). This failure is well before any thread library issues are encountered. I ran a test of mmap on FreeBSD current to see what address space I was able to mmap. Using objdump to identify the pages that the test program was linked to load in, the test program was able to mmap MAP_ANON, MAP_FIXED, MAP_NOCORE all memory from 0x0 to 0xbfc00000, except the pages at which the test program was mapped at. I'll look further into how Wine is mmap'ing memory on FreeBSD. Gerald then wondered where the change needed to be made - Wine or FreeBSD. John further explained: From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: I see isn't documented. From the documentation, I would have expected the Wine code to work - it doesn't seem unreasonable to me. I think the FreeBSD kernel code needs to change. If this is to happen, it will only be in the FreeBSD5 tree. It is too late in the FreeBSD4 branch to make that sort of change since it's not really a bug fix. To get Wine to work on FreeBSD4, there needs to be a way of making the reservation code optional. A simple mmap test in configure which snaffles memory above 0x80000000 and then tries to mmap some more memory without specifying a fixed address would detect if mmap is behaving in a way that would allow the Wine reservation code to function. For FreeBSD5, which will become the stable branch sometime soon, I think the kernel code needs to change. I have a FreeBSD src commit bit, but I'm not a vm person, so I can only prototype a change and submit it for review. I'm not sure if the other developers will regard this change favourably, though. They may take the attitude that if Wine can be made to work with the FreeBSD kernel code as is, then Wine should be coded accordingly. If the Wine code was restructured to make the reservation code optional, that would cover both FreeBSD4 and FreeBSD5. Then, if the FreeBSD mmap algorithm was to change in the future, the build could start using the reservation code at that time. Alexandre didn't want that fix for FreeBSD4, Well, there's a reason for that reservation code, and it's that some Windows apps require it; so unless you find some other way to ensure that FreeBSD never allocates anything above 0x80000000, the reservation code can't really be made optional. So... that was the status last month and no one has mailed the list to report any updates since then. If you're on FreeBSD 4.9, things will break if you upgrade to 4.10. If you're on 5.2 things are broken. From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: the 2.x series they (re)started from the BSD4-lite code. Perhaps the most notable caracteristic of FreeBSD is it's unified VM/cache; mostly the masterpiece of John Dyson. I understand NetBSD was more conservative during this particular stage. About books.. there is "The Design and Implementation of FreeBSD", by Kirk McKusick and others. DragonFly and FreeBSD-5.x have a lot of divergences though, and I doubt they are covered in any book. hope that helps, Pedro. Alex Burke wrote: Hi, I have been reading up allot on the history of BSD based operating systems, and I know that Net/FreeBSDs came from 386BSD 0.1 (based heavily on 4.3 Net/2 tape with missing files replaced) with patch kits applied. As I understand it, 386BSD 1.0 was the continuation of 0.1 but with the kernel modularized. Since I have read that DragonFlyBSD aims to eventually try to run large components (such as VFS) in user land, I was wondering whether conceptually they are at all similar? What is the eventual aim of DragonFlyBSD and userland services? I realise that this might seem stupid question since the code has changed so much (4.4BSD-Lite code integrated, VM system...etc) but I guess this is my attempt to try to understand things better, hopefully with the aim of one day being able to help in some form. I also wanted to ask if anybody knew any good books about the architecture of BSD systems that can be read by someone no already knowledgeable about kernel design. Thanks in advance, Alex J Burke From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: """ Will DragonFly use a dynamic /dev filesystem, as in devfs? Current plans are to keep the existing filesystem model, with the removal of minor/major numbering. There may be a 'devd' process to handle dynamic devices. There are other features to complete first before this is tackled. """ Does not say it is not preferred, it does say that of the time of that writing there where more important things then devfs. -- mph From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: of O(d), where d is the depth of the shadow tree. But is it worth for any effort and extending the namecache struct further? I doubt if ever anyone would want to create an overlay of about hundred levels... even ten would be pretty eccentric. not at all. i'm thinking of packaging systems which use layered filesystems to show packages (or not), stuff like this... now, if we want to e.g. rename the whole group, we just start at ncp_shadowroot and cycle through ncp_shadowlinks. Renaming seems to be a truly individual act -- at least, I can't think of any useable semantics for an operation like "group rename". i expressed myself wrongly. i ment "rename which requires us to rename all associated namecache shadow entries in the group". AFAICS the critical one is breaking down the parent/child relation. My idea is that a p/c cut should imply breaking down shadower/shadowed relatons. If we do that, the upper layer will see the lack of the link into the lower one, and will recreate that as it's appropriate in the new situation. The upper layer wouldn't be aware of the rename event in the lower one as such. I don't understand this parent/child thing. could you please elaborate on this? We don't have disconnected namecache entries, but maybe it's something else you are talking about. the only thing that (still) bothers me is: what happens if somebody locks (cache_lookup) an unresolved (and thus unconnected) nullncp and locks the shadowed ncp (where the system doesn't know about yet about this connection). imagine two processes doing this, each in different order: time thread1 thread2 1 lock null lock lower 2 resolve null (blocks) lock null (blocks) we have a deadlock. resolving null means having to lock lower, but this fails as another thread already obtained the lock. Locking is a group operation. Therefore it must be group atomic -- eg., locking the group can't mean obtaining a lock (in the old sense) on each entry simultaneously. Yes, unless the null layer isn't yet associated with the lower layer. There must be one distinguished parameter in the group which carries locked/unlocked state. In my approach it's the nc_exlocks field of the group head. If a thread has that bumped, it has obtained the lock on the whole group. yes, all roger. Regarding your example, thread2 won't perform a separate "lock null" step. unless it just wants to lock two entries of which it doesn't know the shadow assocition. maybe for renaming or whatnot. cheers simon -- Serve - BSD +++ RENT this banner advert +++ ASCII Ribbon /"\ Work - Mac +++ space for low ????????? NOW!1 +++ Campaign \ / Party Enjoy Relax | http://dragonflybsd.org Against HTML \ Dude 2c 2 the max ! http://golden-apple.biz Mail + News / \ Attachment: PGP.sig --Boundary..85347.1348360655.multipart/signed Content-Type: application/octet-stream; name="pgp00003.pgp" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="pgp00003.pgp" Content-Description: "Description: This is a digitally signed message part" LS0tLS1CRUdJTiBQR1AgU0lHTkFUVVJFLS0tLS0KVmVyc2lvbjogR251UEcg djEuNC4xIChEYXJ3aW4pCgppRDhEQlFGRDFwUFVyNVMrZGs2ejg1b1JBdmg2 QUowUUxjeERnVDhsWGdKMXdyQ0xTdHU0NW5zQ1JRQ2dneTdHClc5Q3pMOGhB QjdEUitKQ3hnVnhScG4wPQo9Q2s3VwotLS0tLUVORCBQR1AgU0lHTkFUVVJF LS0tLS0K --Boundary..85347.1348360655.multipart/signed-- From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: there. MFS_ROOT and MFS_ROOT_SIZE are old aliases for the MD_* counterparts. The full patch (including some LINT additions and minor fixes) is at: http://leaf.dragonflybsd.org/~swildner/kopts.diff I'll post a heads up to users@ when I commit it. Sascha -- http://yoyodyne.ath.cx From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: an issue with re(4) TX checksum hardware offloading. Interesting, I didn't have this problem on my 8169S(PCI) and it seems to only plague PCIe devices. Can you help me to do a test? on non-re(4) side: tcpdump -nv -i on your re(4) side: ping -n -c 1 -s 1512 ping -n -c 10 -s 33 Give me the output of the tcpdump. Thanks Best Regards, sephe -- Live Free or Die From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: The organization should choose a single administrator to submit its application via the GSoC web app between March 5-12, 2007. It would be very nice if DF was in it this year. Indeed, I'm happy to help with that process. I work for Google and acted as a mentor for the bzr project last year, I'm happy to do that for DragonFly this year if that's helpful. From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: 'consumer' OS's (Windows, MacOS "classic", OS-X) have some kind of special scheduler hooks or API's for these kinds of scenarios (DirectSound, CoreAudio, etc) and that this is a 'hard problem' (TM) It is a hard problem (thats my impression from reading the Linux audio mailing lists). If I were a musician or pro audio user, I would buy an Apple only for CoreAudio, although Linux is now a much better audio platform than it was some years ago. If you want to run pro audio software (on BSD) you want to run a jack server. Has anyone tried to port jackd or jackdmp to Dragonfly? Is there any chance that it can run low-latency (64 samples / 48khz = 1.3 ms)? From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: Normally if the CHS fields for a partition are set to all 1's if they would otherwise wrap. This typically causes BIOSes to properly detect that the disk should be put in Large mode. If this option is specified the CHS fields will be set to wrapped values instead. This option may be needed on very old PCs. And from the fact that you have to open up your computer to boot from CD I'd say that you computer is quite old. PS: the first sentence might need some work, as a non-native speaker it took me some time to figure it out. -- Erik Wikstr?m From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: A) probably single, or soon will be and B) don't sleep much anyway! ;-) Looking forward to a 'test drive'... Bill Hacker From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: dominant-carrier telcos) one guess is that you are 'grandfathered' on a service package they no longer offer at all. Hence they *want* those still on it to cancel and upgrade. Denying a proper RR becomes a tool in that progression, will probably be justified on the 'obsolete package, no-such feature, end of story', grounds. BTW: the IP address (63.195.85.27) is not in any "DNS based blocklist" that I know of. It's not even "classified" as dynamic IP in any of those. Moreover, there is a PTR record for it (as those who claim to know something about RFCs could have easily checked.) ACK - but that is part of the burden you need to get out from under. Within a dynamic block or not, (SORBS don't list it as such...) that RR is precisely the sort of RR that *specifies* a dynamic IP not assigned to any one organization: adsl-63-195-85-27.dsl.snfc21.pacbell.net rDNS quite aside, we would catch it in three different acl's on the 'dsl' string match on my servers. Dunno how Matt is doing it, but suspect a similar tool. It would be nice if it was possible to configure sendmail to not block any STARTTLS secure mail regardless of the ip or rDNS of the sender, That's not a good idea. Spammers can easily set up TLS. Not a good idea 'naked' - but while RFC provides wiggle-room w/r whether one uses TLS, an SSL tunnel, VPN, matching certs, or whatever - RFC and BCP are quite specific that the relay-submission function (as used by customer's MUA's) - require valid authentication. Just for comparison, Exim's flag is tested with simply: authenticated = * where the right side *could* be a complex test or a lookup. But we've already matched to a specific port AND the non-standard protocol our user community must utilize as well as their UID:GID and pwd match when we set that flag. I'm sure sendmail's rule syntax is different, but trust that the same functionality already exists. ;-) as you web-page suggests; but to my knowledge, such configuration of sendmail is quite non-trivial, so most people don't use it. If you could provide some examples on the web-page where you make this suggestion, or, better yet, include such examples in the default configuration file, it would, IMHO, be the best approach to this problem. I'll take a look, thanks for the suggestion. Bill From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: but the link is not changed when you upgrade e.g. from 1.10 to 1.12. Sascha -- http://yoyodyne.ath.cx From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: This can't speed up the build, however it can prevent it from stalling. For 2.0 we really need to have working packages on release day. -Matt Matthew Dillon From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: *chronically* not attached to the same MB, cable/SATA/USB/FW as the last time used... . ..BSD bootblocks+Gag I like. FICL I can actually understand the inner workings of - but not the *use* of. Too seldom required to use it to remember *how*. And the so-called 'advanced' OS X 'Openfirmware' or Grub, I would throw rocks at were they not so adept at hiding behind an expensive LCD screen.... Keep it simple - let the complexity be chained into.. Bill From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: What can you do? You can compile everything with debugging information and obtain core files of the crashing apps. Then you can (try to) analyse the core files with gdb. To compile the base system with full debugging information you put the following in /etc/make.conf: CFLAGS+=-g STRIP= To build pkgsrc packages with full debugging information you put the following into /usr/pkg/etc/mk.conf: CFLAGS+=-g INSTALL_UNSTRIPPED=yes You can also use ktrace in your debugging. From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: My free time on summer. As graduate student I've got some duties at the university. On April I'm going to two CTF competitions (one on 2-4 of April in Moscow and one on 23-26 of April in Ekaterinburg). On may there will be some teaching activities (exams for the undergraduates) but they shouldn't take much time. On this summer I've got almost free of duties may, june and august. On july I'll be spending a weak for summer school where I'll be teaching students some basic hacking techniques. Also I'll go to my parents for a weak or a weak and a half. The rest of my time I can spend for the project. Wheew... Thanks for reading that far! I hope, its not too late for me to write this letter :) -- Best regards, Dmitry A. Stephantsov From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: ls -l /pfs/data/yy -rw-r--r-- 1 root wheel 78 Mar 31 22:22 /pfs/data/yy From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: ls -l /pfs/data/yy -rw-r--r-- 1 root wheel 78 Mar 30 22:03 /pfs/data/yy Also on the remote slave system I have the /pfs/data mounted as /data but the data in the mount does not seem to change unless I umount and re mount it. From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: ls -l /data/yy /pfs/data/yy -rw-r--r-- 1 root wheel 39 Mar 30 22:03 /data/yy -rw-r--r-- 1 root wheel 78 Mar 30 22:03 /pfs/data/yy Thanks, Dylan From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: May 15 15:53:37 backup_a kernel: ad6: TIMEOUT - READ_DMA retrying (1 retry left) LBA=49814018 May 15 15:53:38 backup_a kernel: ad6: WARNING - READ_DMA soft error (ECC corrected) LBA=49871650 May 15 15:53:38 backup_a kernel: ad6: WARNING - READ_DMA UDMA ICRC error (retrying request) LBA=49871682 May 15 15:53:38 backup_a kernel: ad6: timeout waiting to issue command May 15 15:53:38 backup_a kernel: ad6: error issuing READ_DMA command May 15 15:53:39 backup_a kernel: ad6: WARNING - READ_DMA UDMA ICRC error (retrying request) LBA=49871682 May 15 15:53:39 backup_a kernel: ad6: FAILURE - READ_DMA status=5 error=61 LBA=49871682 May 15 15:53:39 backup_a kernel: ad6: WARNING - READ_DMA soft error (ECC corrected) LBA=12671810 May 15 15:53:39 backup_a kernel: ad6: WARNING - READ_DMA soft error (ECC corrected) LBA=12671842 May 15 15:53:39 backup_a kernel: ad6: WARNING - READ_DMA soft error (ECC corrected) LBA=12671906 May 15 15:55:42 backup_a syslogd: kernel boot file is /boot/kernel May 15 15:55:42 backup_a kernel: ad6: FAILURE - device detached May 15 15:55:42 backup_a kernel: subdisk6: detached May 15 15:55:42 backup_a kernel: ad6: detached Does this mean the controller is reporting the disk is gone? Dylan From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: May 15 15:53:37 backup_a kernel: ad6: TIMEOUT - READ_DMA retrying (1 retry left) LBA=49814018 May 15 15:53:38 backup_a kernel: ad6: WARNING - READ_DMA soft error (ECC corrected) LBA=49871650 May 15 15:53:38 backup_a kernel: ad6: WARNING - READ_DMA UDMA ICRC error (retrying request) LBA=49871682 May 15 15:53:38 backup_a kernel: ad6: timeout waiting to issue command May 15 15:53:38 backup_a kernel: ad6: error issuing READ_DMA command May 15 15:53:39 backup_a kernel: ad6: WARNING - READ_DMA UDMA ICRC error (retrying request) LBA=49871682 May 15 15:53:39 backup_a kernel: ad6: FAILURE - READ_DMA status=5 error=61 LBA=49871682 May 15 15:53:39 backup_a kernel: ad6: WARNING - READ_DMA soft error (ECC corrected) LBA=12671810 May 15 15:53:39 backup_a kernel: ad6: WARNING - READ_DMA soft error (ECC corrected) LBA=12671842 May 15 15:53:39 backup_a kernel: ad6: WARNING - READ_DMA soft error (ECC corrected) LBA=12671906 May 15 15:55:42 backup_a syslogd: kernel boot file is /boot/kernel May 15 15:55:42 backup_a kernel: ad6: FAILURE - device detached May 15 15:55:42 backup_a kernel: subdisk6: detached May 15 15:55:42 backup_a kernel: ad6: detached Does this mean the controller is reporting the disk is gone? Dylan I got similar messages using two disks with an old WD sata disk (among the first 250 GB disks I think) and nata during boot. Sometimes the disk attached, but mostly not. It attached ok in OpenBSD and NetBSD but it took quite some time there to get done (error handling?). However, as far as I can remember, it worked ok in DragonFly as long as it attached during boot. I don't think came up with a good solution. Using a different disk or an AHCI controller is probably easiest ;-) Keep digging though, Max From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: -R Turn on the TRIM enable flag. If enabled, and if the underlying device supports the BIO_DELETE command, the file system will send a delete request to the underlying device for each freed block. The trim enable flag is typically set when the underlying device uses flash-memory as the device can use the delete command to pre-zero or at least avoid copying blocks that have been deleted. -E Use TRIM to erase the file system before creating it. If used and the underlying device supports TRIM, the device's LBAs will be erased using the TRIM command before the file system is cre- ated. From bogus@does.not.exist.com Thu May 5 13:23:34 2016 From: bogus@does.not.exist.com () Date: Thu, 05 May 2016 20:23:34 -0000 Subject: No subject Message-ID: -E TRIM Device(s) before FS creation. Should only be called if all devices are SSDs that support TRIM. Any suggestions, ideas, or feedback is welcome. Thanks, Tim