[osol-discuss] Re: Re: Re: Fire!! core dumped!
Frank,
How did you think it's calling pthread_mutex_lock(NULL), but no other e.g.
pthread_mutex_lock("invalid address, e.g. 0xf)?
I think the NULL pointer into pthread_mutex_lock is not accordant with signal
SEGV (access to address exceeded protections).
BR,
Eric
This message posted from opensolaris.org
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Re: [osol-discuss] partition tables/slices
>
>> So what happens if I pull out a drive from my SPARC based
>> solaris 8 system and plug it into a intel box...
>
>
> Solaris-SPARC uses a different disk TOC than Solaris-x86.
>
here is a source bit that seems to indicate where the difference lay.
from /usr/src/cmd/fmthard/fmthard.c slightly reformatted by me :
int main(int argc, char **argv) {
intfd, c;
char *dfile;
char *vname;
struct stat statbuf;
#if defined(_SUNOS_VTOC_8)
struct vtoc disk_vtoc;
#endif /* defined(_SUNOS_VTOC_8) */
struct dk_gpt *disk_efi;
struct dk_geom disk_geom;
int n;
really our options are either _SUNOS_VTOC_8 or _SUNOS_VTOC_16 depending on
if we are going to have 8 slices or 16.
It would be nice to change the Sparc implementation to 16 slices .. but that
would break some things .. badly.
I think that EFI partitions change things again.
Dennis
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Re: [osol-discuss] partition tables/slices
> >> So what happens if I pull out a drive from my SPARC based >> solaris 8 system and plug it into a intel box... > > > Solaris-SPARC uses a different disk TOC than Solaris-x86. > > X86 disks use a disk label that is derived from the PC world, > with Master Boot Records, fdisk, partitions and the like; within > one of these partitions, Solaris-X86 writes out a Solaris-SPARC-like > disk label that spells out the slice info that the rest of the > Solaris tools use. A far more curious and useful bit is the result we get from fmthard in Solaris x86. While both Sparc and x86 implementations of Solaris have both fmthard and prtvtoc you will note that only the x86 side of life allows for 16 slices while Sparc is limited to 8. thus : # format -e Searching for disks...done AVAILABLE DISK SELECTIONS: 0. c1t0d0 /[EMAIL PROTECTED],0/pci9004,[EMAIL PROTECTED]/[EMAIL PROTECTED],0 1. c3t0d0 /[EMAIL PROTECTED],0/pci1011,[EMAIL PROTECTED]/pci1000,[EMAIL PROTECTED]/[EMAIL PROTECTED],0 2. c4t1d0 /[EMAIL PROTECTED],0/pci9004,[EMAIL PROTECTED]/[EMAIL PROTECTED],0 Specify disk (enter its number): Specify disk (enter its number): 2 selecting c4t1d0 [disk formatted] FORMAT MENU: disk - select a disk type - select (define) a disk type partition - select (define) a partition table current- describe the current disk format - format and analyze the disk fdisk - run the fdisk program repair - repair a defective sector label - write label to the disk analyze- surface analysis defect - defect list management backup - search for backup labels verify - read and display labels save - save new disk/partition definitions inquiry- show vendor, product and revision scsi - independent SCSI mode selects cache - enable, disable or query SCSI disk cache volname- set 8-character volume name ! - execute , then return quit format> inq Vendor: SEAGATE Product: ST336704LSUN36G Revision: 032C format> format> pa PARTITION MENU: 0 - change `0' partition 1 - change `1' partition 2 - change `2' partition 3 - change `3' partition 4 - change `4' partition 5 - change `5' partition 6 - change `6' partition 7 - change `7' partition 9 - change `9' partition select - select a predefined table modify - modify a predefined partition table name - name the current table print - display the current table label - write partition map and label to the disk ! - execute , then return quit partition> pr Current partition table (original): Total disk cylinders available: 4424 + 2 (reserved cylinders) Part TagFlag CylindersSizeBlocks 0 unassignedwm 1 - 4423 33.88GB(4423/0/0) 71055495 1 unassignedwu 0 0 (0/0/0) 0 2 backupwu 0 - 4423 33.89GB(4424/0/0) 71071560 3 unassignedwu 0 0 (0/0/0) 0 4 unassignedwu 0 0 (0/0/0) 0 5 unassignedwu 0 0 (0/0/0) 0 6 unassignedwu 0 0 (0/0/0) 0 7 unassignedwu 0 0 (0/0/0) 0 8 bootwu 0 -07.84MB(1/0/0) 16065 9 unassignedwu 0 0 (0/0/0) 0 partition> and prtvtoc reveals what format does not : # prtvtoc /dev/rdsk/c4t1d0s0 * /dev/rdsk/c4t1d0s0 partition map * * Dimensions: * 512 bytes/sector * 63 sectors/track * 255 tracks/cylinder * 16065 sectors/cylinder *4426 cylinders *4424 accessible cylinders * * Flags: * 1: unmountable * 10: read-only * * First SectorLast * Partition Tag FlagsSector CountSector Mount Directory 0 000 16065 71055495 71071559 2 501 0 71071560 71071559 8 101 0 16065 16064 10 000 16065 1606500 1622564 13 0001622565 6426000 8048564 # there are slices 10 and 13 there. In fact, you can go up to number 15 for a total of 16 slices. Perfectly valid and with the newer 500GB disks they may be useful. Dennis ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] partition tables/slices
So what happens if I pull out a drive from my SPARC based solaris 8 system and plug it into a intel box... Solaris-SPARC uses a different disk TOC than Solaris-x86. X86 disks use a disk label that is derived from the PC world, with Master Boot Records, fdisk, partitions and the like; within one of these partitions, Solaris-X86 writes out a Solaris-SPARC-like disk label that spells out the slice info that the rest of the Solaris tools use. On a Solaris-SPARC system, there is no fdisk-managed MBR, partition table and the like - only a disk label full of slice info. There may be (I haven't tried it) code in the Solaris disk driver that can deal with this difference, so I can't tell you what would happen if you actually tried it. -John ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] partition tables/slices
Hello, I'm in a muddle over how solaris slices relate to the regular x86 partitions? Basically my understanding is that a single x86 disk partition gets mapped as a single drive in solaris, and within that partition you can chop it up into normal 8 slices. So what happens if I pull out a drive from my SPARC based solaris 8 system and plug it into a intel box, will a Linux OS see this as a single partition for a Sun sparc architecture, if so then how could a linux system mount an individual solaris slice holding a ufs filesystem? Can OpenSolaris 10 cope with the various partitionings that a multi boot linux/windows machine might have? with the awful mess of four primary partitions, with embedded extra partitions in a nonimnated extended partition. How does this all work across platforms? confused of london This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] whats broken in b56
Hello, A few questions from a old sparc solaris 8 admin. I was at the london TechDays ten days ago, I picked up a copy of OpenSolaris 10 b56, and I've attempted to install this on several machines all seem to fail, here's a quick of the problems I've encountered. HP DL380G5 the boot up basically crashed because it didn't understand the SmartArray P400 chip Isn't this a basic mid to high end x86 server An old HP e8000 server, boy the solaris bootup basically just looped forever probeing it's scsi buses I never got beyond device probeing!!! On a vmware server running on a motherboard with Pentium III chips, hey!!! it actuallty managed to install, only I'm stuck with this graphics login rubbish, am I really forced to run this desktop X shit!! Sorry, I'm bascially a server guy the graphics stuff is for the workstation/developer geeks, why am I being forced to run it on a server surely we should just leave the "pretty pretty" stuff to the web/windows/mac peeps!!! Sorry for the rant but all this focus on the desktop stuff for me is a major misfocus, basically I just want to see how solaris10 and opensolaris are as server OS, why try to lock horns with the linux/apple/windows desktop crowd? And why assume that anyone who install on a x86 platform must runa X desktop, am I the last command line person? Again apologies for people who can't live without a all singing all dancing graphical desktop, it just leaves me completely cold, especially when my main reaction is to just turn it off and walk away. This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] [n00b] Solaris Install Help (No, it's not in the manual!)
> Hy there! > I have a little problem installing Solaris 10 on my system. > I'm a complete n00b when it comes to Solaris,but I'm a fast learner so > please bare with me. > I got the ISO from the Solaris website.I burn it.I follow and follow the > install steps until it asks me from what do I wish to install Solaris 10 > from.I select CD\DVD.It gets to 5% then it says something like " Solaris > instalation was not found on the selected medium".Then what the f... did it > boot from?! A floppy?! > Ok,ok,calm down...maybe there was an error when I burned the DVD.Maybe the > DVD itself had scratches&stuff.So I get another DVD and burn the ISO > again.And again,the same 5% install error. that is bizarre .. so you boot the DVD ? then start the installation configuration steps, like hostname and network and other things like that .. then .. you get this question and everything just stops for you because the DVD is no longer really ummm? seen? detected? that is just bizarre and I never heard of such a thing. drop into the IRC channel and ask around irc://irc.freenode.net/opensolaris Dennis ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] [n00b] Solaris Install Help (No, it's not in the manual!)
Hy there! I have a little problem installing Solaris 10 on my system. I'm a complete n00b when it comes to Solaris,but I'm a fast learner so please bare with me. I got the ISO from the Solaris website.I burn it.I follow and follow the install steps until it asks me from what do I wish to install Solaris 10 from.I select CD\DVD.It gets to 5% then it says something like " Solaris instalation was not found on the selected medium".Then what the f... did it boot from?! A floppy?! Ok,ok,calm down...maybe there was an error when I burned the DVD.Maybe the DVD itself had scratches&stuff.So I get another DVD and burn the ISO again.And again,the same 5% install error. My system is composed of the following : AMD Athlon 2600+ CPU @ 1.91GHz,Asus A7V600 X motherboard,Ati Radeon X1550 AGP 8x video card,Western Digital 120GB WD1200JB SATA HDD,Nec 3540A DVD RW drive,LG CD RW drive Thought maybe it's some hardware issue,that's why I listed my components above. I'm lost at this stage. Thanks in advance. Love&Peace! This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Re: Re: Re: Re: Update to B60 ?
Well it does slightly defeat the purpose of an automatic link if nobody knows what it is. To be honest, I thought the one on the download page *was* the automatically updated version until I saw this thread. Andrew. This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Re: Re: Update to B60 ?
> Why not make the one on the download page the automatically updated one??? A simple script run nightly would take care of that eh ? makes sense to do I would think. dc ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Re: Re: Re: Update to B60 ?
Why not make the one on the download page the automatically updated one??? Andrew. This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Re: Update to B60 ?
> On Sun, Mar 25, 2007 at 04:55:29AM -0700, Andrew Pattison wrote: >> > Incorrect. >> > The site just sometimes doesn't update properly for >> > some reason. Latest is always here: >> > http://opensolaris.org/sxce_dvd >> >> Why is this "magic" URL up-to-date, yet the one on the download page >> (http://www.opensolaris.org/os/downloads/sol_ex_dvd/) still points to >> build 59? > > I believe that one is manually updated by Derek. > The one at http://opensolaris.org/sxce_dvd periodically polls the SDLC > to see if a new build has been released and auto-updates accordingly. The complete URL seems to be thus : http://javashoplm.sun.com/ECom/docs/Welcome.jsp?StoreId=7&PartDetailId=Sol-Express_b60-DVD-x86-SP-G-B&TransactionId=try However you can change that to various versions simply by making the 60 into 55b or whatever. Dennis ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Re: Update to B60 ?
On Sun, Mar 25, 2007 at 04:55:29AM -0700, Andrew Pattison wrote: > > Incorrect. > > The site just sometimes doesn't update properly for > > some reason. Latest is always here: > > http://opensolaris.org/sxce_dvd > > Why is this "magic" URL up-to-date, yet the one on the download page > (http://www.opensolaris.org/os/downloads/sol_ex_dvd/) still points to build > 59? I believe that one is manually updated by Derek. The one at http://opensolaris.org/sxce_dvd periodically polls the SDLC to see if a new build has been released and auto-updates accordingly. -steve -- stephen lau // [EMAIL PROTECTED] | 650.786.0845 | http://whacked.net opensolaris // solaris kernel development ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Containers in practice – pleas
That's exactly the model http://joyent.com/ are using with their accelerators. Seems to be very well recieved by their hosting customers (of which I'm one). On 25/03/07, Adam Luzecki <[EMAIL PROTECTED]> wrote: > One of the primary benefits of Containers is > delegated administration. > Think of a company providing hosting services. > Instead of providing each > customer their own server, they can be provided with > a Solaris > container. Many customers can share resources on the > same set of > servers. The resource management capabilities of > Containers allows the > hosting provider to assign resources to a container > based on an > agreed-upon SLA. Perhaps premium-rate customers get > more resources than > standard-rate customers. That's great example! And the idea can be shown even on Ultra 10! I could design a project based on midrange server, and show it in low scale on my workstation! I will think about something like this. Thanks! This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org -- Rasputin :: Jack of All Trades - Master of Nuns http://number9.hellooperator.net/ ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Re: Re: Update to B60 ?
> Incorrect. > The site just sometimes doesn't update properly for > some reason. Latest is always here: > http://opensolaris.org/sxce_dvd Why is this "magic" URL up-to-date, yet the one on the download page (http://www.opensolaris.org/os/downloads/sol_ex_dvd/) still points to build 59? Cheers Andrew. This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Proposal to include dirfd(3C) into OpenSolaris
[EMAIL PROTECTED] wrote:
> >Do you mean:
> >
> > sprintf(buf, "/proc/self/fd/%d/%s", dfd, dname);
> > dp = opendir(buf);
>
> Yes.
>
> >> Not sure why you have all the dup's in there.
> >
> >Because closedir() will also close the fd passed to fdopendir()
> >and I need the fd for a reliable chdir("..") later.
>
> You can do the chdir before the closedir.
Depends... I am not sure whether this will result in clean code
for libfind.
> The current rm source keeps a stack of open directories but, since it
> must allow for unlimited depth trees it verifies it can go back up before
> closing the parent directory. And it uses only one filedescriptor per
> level (including the one used for opendir)
You may have a look at libfind (e.g. from a recent cdrtools
ftp://ftp.berlios.de/pub/cdrecord/alpha/) to see how to do this without the
need to use filedescriptors. A library cannot eat as many fd's as it may like...
I will still have libfind to use an openat() based implementation
in order to support infinite path length in a thread safe way in case the OS
supports openat(). The basic algorithm will stay the same as today.
Jörg
--
EMail:[EMAIL PROTECTED] (home) Jörg Schilling D-13353 Berlin
[EMAIL PROTECTED](uni)
[EMAIL PROTECTED] (work) Blog: http://schily.blogspot.com/
URL: http://cdrecord.berlios.de/old/private/ ftp://ftp.berlios.de/pub/schily
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
Re: [osol-discuss] Re: Proposal to include dirfd(3C) into OpenSolaris
>[EMAIL PROTECTED] wrote:
>
>>
>> >Linux started to support openat() recently and for this reason,
>> >the best way of implementing a portable way of directory search without
>> >path length limitation is to use:
>> >
>> >ndfd = openat(dfd, "dname", O_RDONLY);
>> >close(dfd);
>> >dfd = ndfd;
>> >dp = fdopendir(dup(dfd));
>>
>> Interesting is, of course, that /proc/self/fd has made this obsolete
>> in an interesting sense.
>>
>> (openat maps to:
>> sprintf(buf, "/proc/self/%d/%s", dfd, dname);
>
>Do you mean:
>
> sprintf(buf, "/proc/self/fd/%d/%s", dfd, dname);
> dp = opendir(buf);
Yes.
>> Not sure why you have all the dup's in there.
>
>Because closedir() will also close the fd passed to fdopendir()
>and I need the fd for a reliable chdir("..") later.
You can do the chdir before the closedir.
The current rm source keeps a stack of open directories but, since it
must allow for unlimited depth trees it verifies it can go back up before
closing the parent directory. And it uses only one filedescriptor per
level (including the one used for opendir)
Casper
___
opensolaris-discuss mailing list
opensolaris-discuss@opensolaris.org
[osol-discuss] Re: Why a GUI installer? - was Re: Re: RE: Solaris on Intel's Classmate P
> > >> but I admit I don't know the technical reasons why > we have > >> the requirements we have today to begin with. > > > > Because responsible folks don't listen to Moinak > Ghosh's Belenix > > minimize_Grub_bootImage - approach? > > The primary requirements are internationalization > (only easy for > western alphabets without a gui) and accessibility > (significantly > more difficult to near-impossible in a non-gui > environment) > > In order to sell a product to the US govt, companies > have to follow > the Americans with Disabilities Act (ADA), which > pulls in the section > 508 statute: > > http://www.section508.gov/ Are you implying that the installer is bloated because it has to support people with disabilities? At least, that's what it seems to me that you're implying, which doesn't make any sense. Slightly off topic, but since you mentioned the disabilities act, doesn't the U.S. government use Linux as well? And I haven't noticed that the RedHat installer provides anything special for people with disabilities. So how's come RedHat gets their foot in the door and doesn't get busted by the federal government? Nevertheless, that all does not explain why the installer has to be bloatware, and why it would have to be in Java. This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
[osol-discuss] Re: Containers in practice – pleas
> One of the primary benefits of Containers is > delegated administration. > Think of a company providing hosting services. > Instead of providing each > customer their own server, they can be provided with > a Solaris > container. Many customers can share resources on the > same set of > servers. The resource management capabilities of > Containers allows the > hosting provider to assign resources to a container > based on an > agreed-upon SLA. Perhaps premium-rate customers get > more resources than > standard-rate customers. That's great example! And the idea can be shown even on Ultra 10! I could design a project based on midrange server, and show it in low scale on my workstation! I will think about something like this. Thanks! This message posted from opensolaris.org ___ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org
