http://blog.gmane.org/gmane.os.netbsd.devel.kernel hourly 1 1901-01-01T00:00+00:00 http://gmane.org/img/gmane-25t.png http://gmane.org http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42133 <pre>It seems that I have to update my understanding of raw and block devices for discs. Using a (non-recent) 6.0_BETA INSTALL kernel and an ST9146853SS 15k SAS disc behind an LSI SAS 1068E (i.e. mpt(4)), I did a dd if=/dev/zero od=/dev/[r]sd0b bs=nn, count=xxx. For the raw device, the troughput dramatically increased with the block size: Block size16k64k256k1M Troughput (MByte/s)41549112 For the block device, throughput was around 81MByte/s independent of block size. This surprised me in two ways: 1. I would have expected the raw device to outperform the block devices with not too small block sizes. 2. I would have expected inceasing the block size above MAXPHYS not improving the performance. So obviously, my understanding is wrong. I then build a RAID 1 with SectorsPerSU=128 (e.g. a 64k stripe size) on two of these discs, and, after the parity initialisation was complete, wrote to [r]raid0b. On the raw device, throghput ranged from 4MByte/s to 97MByte/s depending on bs. On the block devic</pre> Edgar Fuß 2012-05-24T16:26:45 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42127 <pre>Hello, This is a new resource limit to prevent users from exhausting kernel resources that lwps use. - The limit is per uid - The default is 1024 per user unless the architecture overrides it - The kernel is never prohibited from creating threads - Exceeding the thread limit does not prevent process creation, but it will prevent processes from creating additional threads. So the effective thread limit is nlwp + nproc - The name NTHR was chosen to follow prior art - There could be atomicity issues for setuid and lwp exits - This diff also adds a sysctl kern.uidinfo.* to show the user the uid limits comments? christos Index: kern/init_main.c =================================================================== RCS file: /cvsroot/src/sys/kern/init_main.c,v retrieving revision 1.442 diff -u -p -u -r1.442 init_main.c --- kern/init_main.c19 Feb 2012 21:06:47 -00001.442 +++ kern/init_main.c23 May 2012 23:19:31 -0000 &lt; at &gt;&lt; at &gt; -256,6 +256,7 &lt; at &gt;&lt; at &gt; intcold = 1;/* still working on star struct timespec boottime; </pre> Christos Zoulas 2012-05-23T23:37:19 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42122 <pre>In the regular sparc test runs on qemu the emulated sparc machine only has 32MB of ram. In this setup the /usr/tests/lib/libc/sys/t_mincore test "mincore_resid" fails. If we allow qemu ot provide more memory, the test succeeds. The part of the test that fails in "low" memory environments is: An anonymous mmap of 128 pages size is created: addr = mmap(NULL, npgs * page, PROT_READ | PROT_WRITE, MAP_ANON | MAP_PRIVATE, -1, (off_t)0); Nothing in this mapping is touched, so nothing gets paged in/modified. Via mincore() the test now validates that 0 of the mapped pages are resident. So far, this works. Now a mlockall(MCL_CURRENT|MCL_FUTURE) call is done. The test now expects all pages in the above mapping to be resident and again checks via mincore(). This is the part where the test fails: some of the pages are not resident. I do believe this is not a bug in the sparc pmap. Am I wrong? Is this an MI bug? Or a bogus test? Martin </pre> Martin Husemann 2012-05-23T16:15:32 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42111 <pre>What's the suggested method for breaking out of the emulation directory? I want &lt;EMULDIR&gt;/opt/tivoli/tsm/client/ba/bin/dsm.opt to be a symlink to &lt;no-emulation&gt;/usr/pkg/etc/tsm/dsm.opt. I can achieve this with a considerable amount of ../, but that amount depends on the value of &lt;EMULDIR&gt;, more precisely, it's expanded value, which happens to be /usr/pkg/emul/linux32 in the case in question. I was thinking about a /emul/none -&gt; / symlink. I'm not sure how the behaviour of the emulation sort-of-chroot is defined to be if it, for example, looks up /usr/local/foo/bar and e.g. a "real" /usr/local (or /usr/local/foo) exists, but /usr/local/foo (or /usr/local/foo/bar) doesn't. Is this documented anywhere? Has it change since 4.0? </pre> Edgar Fuß 2012-05-21T13:44:51 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42109 <pre>What does this mean: May 21 15:32:12 trave /netbsd: /emul/linux32/usr/bin/dsmc: bad tag 1: [5 4, 2 4, SuSE PaX] </pre> Edgar Fuß 2012-05-21T13:36:28 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42101 <pre>I was somewhat surprised to learn that with libquota, qv_usage etc. were still in units of what someone called "a constant of nature introduced by DEC", e.g., 512-byte "blocks". Given the fields are 64 bits wide, I would have expected them to be in bytes instead, but there are probably reasons not do do that. So, I would at least expect that to be documented or machine-readable. </pre> Edgar Fuß 2012-05-20T16:10:04 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42096 <pre>Is there an interface for reading (or even writing) another process' ulimits? </pre> Edgar Fuß 2012-05-19T16:51:48 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42082 <pre>I just started to play around with the new quota system. I tried to set quotas for a user that didn't own any files on the file system in question. With an "interactive" edquota, I got edquota: /export/test (ufs/ffs quota v2): : bad format With a edquota -h ... -s ..., I got a zero exit status, a following "interactive" edquota showing what I entered, but repquota not showing anything for that user. When I allocate a file for that user, edquota -h ... -s ... works, but an "interactive" edquota throws the same error and disables that user's quota. Is this expected behaviour? </pre> Edgar Fuß 2012-05-14T15:58:46 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42065 <pre>Hi, I see kernel crashes when playing a sample via a USB audio device: uaudio0 at uhub0 port 2 configuration 1 interface 0 uaudio0: Logitech Logitech USB Headset, rev 1.10/10.13, addr 3 uaudio0: ignored input endpoint of type adaptive uaudio_identify_as: too much tSamFreq: 5 uaudio_identify_as: too much tSamFreq: 5 uaudio0: audio rev 1.00 audio1 at uaudio0: full duplex, playback, capture, independent note the "uaudio_identify_as: too much tSamFreq: 5" lines here. With extra debug printf(), I can see that the attach adds sample rates of 8000 and 11025 before the test added in r1.130 of uaudio.c [*] applies and breaks out of the loop. The problem is that the number of frequencies is still set to the old value (i.e., frequency_type = 5) and when playing any audio, the frequency list is looped over in auconv_rateconv_check_rates(). With extra debug there, I see: auconv_rateconv_check_rates: sample_rate = 44100 frequency[0] = 8000, minrate = 8000, maxrate = 8000 frequency[1] = 8000, min</pre> Julian Coleman 2012-05-11T13:42:11 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42063 <pre>Is there some utility, that, given a file system with real data on it, will compute the optimum file system block size for that data set? If not, is it a reasonable assumption that the space allocated to a file will be the file's data size, rounded up to the fragment size? </pre> Edgar Fuß 2012-05-11T10:21:31 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42055 <pre>Does anyone have some real-world experience with RAIDframe (Level 5) performance vs. stripe size? My impression would be that, with a not to large number of components (4+1, in my case), chances are rather low to spread simultaneous accesses to different physical discs, so the best choice seems the file system's block size. On the other hand, the Carnegie Mellon paper suggests something around 1/2*(througput)*(access time), which would amount to more than a Megabyte with my discs. So, what are the real-world benefits of a large stripe size? My application are home and mail directories exported via NFS. </pre> Edgar Fuß 2012-05-10T15:46:38 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42051 <pre>Does anybody understand why this happens on several archs in -current? http://releng.netbsd.org/builds/HEAD/201205041410Z/amiga.build.failed dependall ===&gt; sys/modules/hfs WARNING: ioconf is an experimental feature WARNING: pseudo-root is an experimental feature dependall ===&gt; sys/modules/iic WARNING: ioconf is an experimental feature /home/builds/ab/HEAD/src/sys/dev/i2c/files.i2c:3: file system or option `I2C_SCAN' already defined /home/builds/ab/HEAD/src/sys/dev/i2c/files.i2c:4: attribute `i2cbus' already defined /home/builds/ab/HEAD/src/sys/dev/i2c/files.i2c:5: attribute `i2cexec' already defined /home/builds/ab/HEAD/src/sys/dev/i2c/files.i2c:7: redefinition of `iic' /home/builds/ab/HEAD/src/sys/dev/i2c/files.i2c:8: redefinition of `iic' /home/builds/ab/HEAD/src/sys/dev/i2c/files.i2c:9: duplicate file dev/i2c/i2c.c /home/builds/ab/HEAD/src/sys/dev/i2c/files.i2c:9: here is the original definition /home/builds/ab/HEAD/src/sys/dev/i2c/files.i2c:10: duplicate file dev/i2c/i2c_exec.c /home/builds/ab/HEAD/src</pre> Martin Husemann 2012-05-05T09:46:01 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42044 <pre>Hi I recently committeda fix that broke two puffs tests: src/sys/kern/vfs_syscalls.c 1.451-1.454 http://releng.netbsd.org/b5reports/i386/build/2012.04.30.13.47.52/test.html#fail ed-tcs-summary I have trouble figuring what goes wrong. Both in mountfuzz7 and mountfuzz8, it panics "mount: lost mount" in src/sys/kern/vfs_mount.c:mount_checkdirs() The offending change is below. Basically, it moved from the situation where mount_checkdirs() founds v_usecount ==1 (and it just returns) to the situation where v_usecount == 2 (it calls VFS_ROOT(olddp-&gt;v_mountedhere, &amp;newdp) and panics) Any suggestion on how to fix that proprely? I call vref(vp) before mount_domount() sothat I can acccess vp-&gt;v_mountedhere after mount_domount sets vp to NULL, but it seems this is not The Right Way. Index: vfs_syscalls.c =================================================================== RCS file: /cvsroot/src/sys/kern/vfs_syscalls.c,v retrieving revision 1.451 retrieving revision 1.454 diff -U4 -r1.451 -r1.454 --- vfs_syscalls.</pre> Emmanuel Dreyfus 2012-05-04T15:12:27 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42035 <pre>I have an application where I want to watch USB devices come and go. I've written code basd on usb(4), and it works - but the devices in question are disks, which show up as, for example, umass0: using SCSI over Bulk-Only scsibus1 at umass0: 2 targets, 1 lun per target sd0 at scsibus1 target 0 lun 0: &lt;Generic, External, 2.10&gt; disk fixed So, I'm wondering if there's some way to watch devices come and go beyond what /dev/usb gives me (which stops with the umass attachment). Even some way to inspect the current device tree would help; I can watch umass attach and then query the tree to see what's underneath it. I have a fuzzy memory of seeing something that looked like device-tree data, but the memory's too fuzzy to be of much use here. kern.drivers appears to be part of what I'd want, but only part; I'd much rather not have to scrape dmesg output. :/ The machine in question is currently at 4.0.1. If this is possible with a more recent version but not with 4.0.1, I might be able to talk its admins into swi</pre> Mouse 2012-05-04T00:01:23 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42023 <pre>While trying to muddle my way through the net80211 module mess, I noticed the following line in sys/cdefs.h: #define __link_set_entry(set, idx) (__link_set_begin(set)[idx]) This is the only place where __link_set_begin is referenced in the entire source tree, and I think that it really should have been #define __link_set_entry(set, idx) (__link_set_start(set)[idx]) Fortunately, there is also nothing else in the source tree that uses the __link_set_entry macro(), so this isn't causing any real pain. But I think it should be corrected before someone does try to use it. Any reason this should not be changed? ------------------------------------------------------------------------- | Paul Goyette | PGP Key fingerprint: | E-mail addresses: | | Customer Service | FA29 0E3B 35AF E8AE 6651 | paul at whooppee.com | | Network Engineer | 0786 F758 55DE 53BA 7731 | pgoyette at juniper.net | | Kernel Developer | | pgoyette at netbsd.org | -----</pre> Paul Goyette 2012-04-30T19:06:27 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42005 <pre>Hi There is a choice to be made about returing ENOATTR or ENODATA when we do not find an extended attribute. Here are the existing behaviors: 1) return ENOATTR, while ENODATA is defined and has other usages: NetBSD and MacOS X 2) return ENODATA, while ENOATTR is not defined: Linux 3) return ENOATTR, while ENODATA is not defined: FreeBSD The problem is that some userland program may test (errno == ENODATA) and get a different semantic on NetBSD and Linux. glusterfs does that, for instance. It seems there is no standard about this. OpenGroup only defines ENODATA, but for another usage (STREAM). In order to get the broader compatibility, I suggest patching our errno.h to define ENOATTR as ENODATA. Opinions? Index: sys/errno.h =================================================================== RCS file: /cvsroot/src/sys/sys/errno.h,v retrieving revision 1.39 diff -U 4 -r1.39 errno.h --- sys/errno.h 31 Oct 2006 00:38:07 -0000 1.39 +++ sys/errno.h 30 Apr 2012 09:17:43 -0000 &lt; at &gt;&lt; at &gt; -164,9 +164,9 &lt; at &gt;&lt; at &gt; #d</pre> Emmanuel Dreyfus 2012-04-30T09:20:42 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/42002 <pre>Hello I hace a filesystem mounted on /foo how do I retreive its struct mount? Using namei_simple_user(), I can get a vnode for /foo but its v_mount is the one for the root filesystem. Looking up /foo/ produces an error. (bad address). How do I cross the mount point and get a vnode for the monted filesystem root? Or should I get the struct mount by another method? </pre> Emmanuel Dreyfus 2012-04-30T00:47:08 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/41996 <pre>Hi, Unless anyone objects, I will remove POOL_DIAGNOSTIC. Since default allocator is pool_cache(9), in post-netbsd-5 it is not really useful nor really used (note that after kmem(9) changes it does not compile). For pool_cache(9), existing consistency checks and safeguards under DEBUG/DIAGNOSTIC are fairly effective. http://www.netbsd.org/~rmind/pool_diag.diff </pre> Mindaugas Rasiukevicius 2012-04-28T23:51:08 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/41992 <pre>Hi On -current and netbsd-6, starting extended attributes is broken. I just fixed a NULL pointer reference that caused a panic, but there is another problem. ufs_extattr_autostart() needs to lookup .attribute/user and .attribute/system directories on the top of the filesystem. It is done through ufs_extattr_lookup(),which does this: memset(&amp;cnp, 0, sizeof(cnp)); cnp.cn_nameiop = LOOKUP; cnp.cn_flags = ISLASTCN | lockparent; cnp.cn_cred = l-&gt;l_cred; cnp.cn_nameptr = pnbuf; error = copystr(dirname, pnbuf, MAXPATHLEN, &amp;cnp.cn_namelen); if (error) { /* removed error handling for clarty sake */ } cnp.cn_namelen--; /* trim nul termination */ vargs.a_desc = NULL; vargs.a_dvp = start_dvp; vargs.a_vpp = &amp;target_vp; vargs.a_cnp = &amp;cnp; error = ufs_lookup(&amp;vargs); ufs_lookup now returns ENOENT while the directory exists. It did not do that on netbsd-5. Any idea of what when wrong? </pre> Emmanuel Dreyfus 2012-04-28T17:36:08 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/41982 <pre>Hello, I bought a new disk which included a warning label along the line of "Warning: Uses 4K sector size!". I plugged it in, and NetBSD identifies it as: umass0 at uhub6 port 1 configuration 1 interface 0 umass0: JMicron USB to ATA/ATAPI bridge, rev 2.00/1.00, addr 2 umass0: using SCSI over Bulk-Only scsibus0 at umass0: 2 targets, 1 lun per target sd0 at scsibus0 target 0 lun 0: &lt;WDC WD75, 00AARX-00N0YB0, &gt; disk fixed sd0: 698 GB, 16383 cyl, 16 head, 63 sec, 4096 bytes/sect x 183143646 sectors disklabel identifies it as: ------------------------------ # /dev/rsd0d: type: SCSI disk: 00AARX-00N0YB0 label: fictitious flags: bytes/sector: 4096 sectors/track: 63 tracks/cylinder: 16 sectors/cylinder: 1008 cylinders: 16383 total sectors: 183143646 rpm: 3600 interleave: 1 trackskew: 0 cylinderskew: 0 headswitch: 0 # microseconds track-to-track seek: 0 # microseconds drivedata: 0 [---] ------------------------------ So far so good; I can add a slice, and run newfs on the slice. However, I</pre> Jan Danielsson 2012-04-28T01:52:38 http://comments.gmane.org/gmane.os.netbsd.devel.kernel/41977 <pre>Folks, I would like to commit the attached patches to modularize the net80211 code. The changes include: 1. Protecting the inclusion of "opt_inet.h" with #ifdef _KERNEL 2. Removing CTLFLAG_PERMANENT from all of the sysctl(8) variables 3. Replacing the use of link_set with direct calls to all of the crypto/auth initializers. At a future time, I propose to split the crypto/auth stuff out into their own modules, but that will require a bit more planning and a lot more testing. :) I have tested the attached changes on my home system, using a modular if_rum(4) driver (which I plan to commit separately). Comments/feedback welcomed... ------------------------------------------------------------------------- | Paul Goyette | PGP Key fingerprint: | E-mail addresses: | | Customer Service | FA29 0E3B 35AF E8AE 6651 | paul at whooppee.com | | Network Engineer | 0786 F758 55DE 53BA 7731 | pgoyette at juniper.net | | Kernel Developer | | pgoyette at netbsd.org </pre> Paul Goyette 2012-04-27T12:51:30 Search the mailing list at Gmane query http://search.gmane.org/?group=$group=gmane.os.netbsd.devel.kernel