2.4.0-test4 possible memory issues on boot (re:2.4.0-test4 no boot?)
Hello, all-
I noticed this thread on linux-kernel, an thought I'd throw my two
cents in. I've cc'ed debian-sparc because I'm beginning to suspect
that, at least in my case, this may be an issue with SIMM placement,
and thus something that some of the good folks there may have run
into. I'm not nearly brave enough at this point to actually
subscribe to linux-kernel, so please cc me on any messages sent to
linux-kernel.
That said...
I am seeing nearly identical behavior on a 2-way Sparc 10. On boot,
it claims to be starting linux and then, after announcing a call to
free_all_bootmem, it freezes solid, and won't respond to stop-A.
Here's /proc/cpuinfo:
cpu : Texas Instruments, Inc. - MicroSparc
fpu : SuperSparc on-chip FPU
promlib : Version 3 Revision 2
prom : 2.12
type : sun4m
ncpus probed : 2
ncpus active : 2
Cpu0Bogo : 39.83
Cpu1Bogo : 39.93
MMU type : TI Viking
invall : 0
invmm : 0
invrnge : 0
invpg : 0
contexts : 65536
CPU0 : online
CPU1 : online
I've compiled (and trimmed) some potentially interesting behavior
from the console on boot into 2.4.0-test4, 2.4.0-test1, and 2.2.15.
In the case of 2.4.0-test4, it simply doesn't boot, but the bootmem
information it gives is slightly different from 2.4.0-test1, and
will hopefully be helpful.
With 2.4.0-test1, it boots, but doesn't seem to find anywhere near
all the memory. In fact, the arithmetic suggests that it's only
dealing with 64MB of memory, instead of 96MB. According to the
info I found on docs.sun.com, I believe I've to the 6x16MB SIMMS in
the correct places, but the order does seem (to me at least) to be
deeply pathological, so I'm a bit suspicious.
With the 2.4.0-test1 kernel, I also dumped the output from
/proc/openprom/memory@0,0/available, in case that it helpful.
Sadly, I was scared away from the /proc/openprom module when I
compiled 2.2.15 by the fact that it was experimental, so I don't
have anything else to compare the 2.4.0-test1 results with.
If there's any other information whic would be helpful, or anything
else I should try, please let me know. I appear not to have saved
a copy of my 2.4.0-test1 config, but I do have the 2.4.0-test4 and
2.2.15 configs available if they would be of interest, but this is
already probably far too long a message.
Thanks-
David
/* Attempt to boot 2.4.0-test4 */
SPARCstation 10 MP (2 X 390Z50), Keyboard Present
ROM Rev. 2.12, 96 MB memory installed, Serial #3170570.
Ethernet address 8:0:20:1b:11:49, Host ID: 7230610a.
Testing Memory -
Initializing Memory |/-\|/-\|/-\|/-\
Boot device: /iommu/sbus/espdma@f,400000/esp@f,800000/sd@3,0 File and args:
SILO boot:
linux 2.4.0-test1 2.4.0-test4
fcs
boot: 2.4.0-test4
PROMLIB: obio_ranges 5
bootmem_init: Scan sp_banks, init_bootmem(spfn[1f1],bpfn[1f1],epfn[c000])
free_bootmem: base[0] size[1000000]
free_bootmem: base[4000000] size[1000000]
free_bootmem: base[8000000] size[1000000]
reserve_bootmem: base[0] size[1f1000]
reserve_bootmem: base[1f1000] size[1800]
init_bootmem: return end_pfn[c000]
Booting Linux...
mem_init: Calling free_all_bootmem().
SPARCstation 10 MP (2 X 390Z50), Keyboard Present
ROM Rev. 2.12, 96 MB memory installed, Serial #3170570.
Ethernet address 8:0:20:1b:11:49, Host ID: 7230610a.
** Power cycle **
/* Boot 2.4.0-test1 */
Testing Memory -
Initializing Memory |/-\|/-\|/-\|/-\
Boot device: /iommu/sbus/espdma@f,400000/esp@f,800000/sd@3,0 File and args:
SILO boot:
linux 2.4.0-test1 2.4.0-test4
fcs
boot: 2.4.0-test1
PROMLIB: obio_ranges 5
bootmem_init: Scan sp_banks, init_bootmem(spfn[1e4],bpfn[1e4],epfn[d000])
free_bootmem: base[0] size[1000000]
free_bootmem: base[4000000] size[1000000]
free_bootmem: base[8000000] size[1000000]
free_bootmem: base[c000000] size[1000000]
reserve_bootmem: base[0] size[1e4000]
reserve_bootmem: base[1e4000] size[1a00]
init_bootmem: return end_pfn[d000]
Booting Linux...
mem_init: Calling free_all_bootmem().
PROMLIB: Sun Boot Prom Version 3 Revision 2
Linux version 2.4.0-test1 (dbutts@sukhoyev) (gcc version 2.95.2 20000313 (Debian GNU/Linux)) #1 SMP Fri Jun 16 13:04:47 EDT 2000
ARCH: SUN4M
TYPE: Sun4m SparcStation10/20
Ethernet address: 8:0:20:1b:11:49
Boot time fixup v1.6. 4/Mar/98 Jakub Jelinek (jj@ultra.linux.cz). Patching kernel for srmmu[TI Viking]/iommu
On node 0 totalpages: 53248
zone(0): 53248 pages.
zone(1): 0 pages.
zone(2): 0 pages.
Found CPU 0 <node=ffd4b120,mid=8>
Found CPU 1 <node=ffd4b3f0,mid=10>
Found 2 CPU prom device tree node(s).
Power off control detected.
Kernel command line: root=/dev/sda1 ro
Calibrating delay loop... 39.83 BogoMIPS
Memory: 58060k available (1288k kernel code, 248k data, 124k init) [f0000000,0d000000]
Dentry-cache hash table entries: 8192 (order: 4, 65536 bytes)
Buffer-cache hash table entries: 1024 (order: 0, 4096 bytes)
Page-cache hash table entries: 16384 (order: 4, 65536 bytes)
kmem_create: Poisoning requested, but con given - bdev_cache
Inode-cache hash table entries: 4096 (order: 3, 32768 bytes)
kmem_create: Poisoning requested, but con given - inode_cache
VFS: Diskquotas version dquot_6.4.0 initialized
POSIX conformance testing by UNIFIX
Entering SMP Mode...
Starting CPU 1 at f018ba28
Calibrating delay loop... 39.94 BogoMIPS
Total of 2 Processors activated (79.77 BogoMIPS).
IOMMU: impl 0 vers 3 page table at f0780000 of size 262144 bytes
<snipped standard boot information>
(root) sukhoyev:~ # free
total used free shared buffers cached
Mem: 58192 16476 41716 0 528 9868
-/+ buffers/cache: 6080 52112
Swap: 131032 0 131032
(root) sukhoyev:~ # cd /proc/openprom/
(root) sukhoyev:/proc/openprom # cd memory\@0,0/
(root) sukhoyev:/proc/openprom/memory@0,0 # ls -la
total 0
dr-xr-xr-x 2 root root 0 Jul 26 00:09 .
dr-xr-xr-x 1 root root 0 Jul 26 00:09 ..
-r--r--r-- 1 root root 0 Jul 26 00:09 .node
-r--r--r-- 1 root root 0 Jul 26 00:09 available
-r--r--r-- 1 root root 0 Jul 26 00:09 name
-r--r--r-- 1 root root 0 Jul 26 00:09 reg
(root) sukhoyev:/proc/openprom/memory@0,0 # cat available
00000000.18f46000.00013000.00000000.18000000.00f45000.00000000.14000000.01000000.00000000.0c000000.01000000.00000000.08000000.01000000.00000000.04000000.01000000.00000000.00000000.01000000
(root) sukhoyev:/proc/openprom/memory@0,0 # sync ; sync ; shutdown -r now
** Reboot **
/* Boot 2.2.15 */
SPARCstation 10 MP (2 X 390Z50), Keyboard Present
ROM Rev. 2.12, 96 MB memory installed, Serial #3170570.
Ethernet address 8:0:20:1b:11:49, Host ID: 7230610a.
Testing Memory -
Initializing Memory
Boot device: /iommu/sbus/espdma@f,400000/esp@f,800000/sd@3,0 File and args:
SILO boot:
linux 2.4.0-test1 2.4.0-test4
fcs
boot: linux
PROMLIB: obio_ranges 5
Booting Linux...
PROMLIB: Sun Boot Prom Version 3 Revision 2
Linux version 2.2.15 (dbutts@sukhoyev) (gcc version 2.95.2 20000313 (Debian GNU/Linux)) #2 SMP Thu Jun 8 22:07:02 EDT 2000
ARCH: SUN4M
TYPE: Sun4m SparcStation10/20
Ethernet address: 8:0:20:1b:11:49
Boot time fixup v1.6. 4/Mar/98 Jakub Jelinek (jj@ultra.linux.cz). Patching kernel for srmmu[TI Viking]/iommu
[0]: v[f0000000,f1000000](1000000) p[00000000]
[1]: v[f1000000,f2000000](1000000) p[04000000]
[2]: v[f2000000,f3000000](1000000) p[08000000]
[3]: v[f3000000,f4000000](1000000) p[0c000000]
[4]: v[f4000000,f5000000](1000000) p[14000000]
[5]: v[f5000000,f5f45000](f45000) p[18000000]
[6]: v[f5f46000,f5f59000](13000) p[18f46000]
Found CPU 0 <node=ffd4b120,mid=8>
Found CPU 1 <node=ffd4b3f0,mid=10>
Found 2 CPU prom device tree node(s).
Power off control detected.
Calibrating delay loop... 39.83 BogoMIPS
Memory: 93604k available (1040k kernel code, 2872k data, 116k init) [f0000000,f6000000]
Dentry hash table entries: 16384 (order 5, 128k)
Buffer cache hash table entries: 131072 (order 7, 512k)
Page cache hash table entries: 32768 (order 5, 128k)
POSIX conformance testing by UNIFIX
Entering SMP Mode...
Starting CPU 1 at f0141980
Calibrating delay loop... 39.94 BogoMIPS
Total of 2 Processors activated (79.77 BogoMIPS).
IOMMU: impl 0 vers 3 page table at f5e40000 of size 262144 bytes
On Sat, Jul 16 Jul 2000 23:24:29 +0200, ragnar@mcaula.net wrote:
> On Sat, Jul 15, 2000 at 09:27:34PM -0600, Robert L. Harris wrote:
> >
> >
> > I just compiled 2.4.0-test4. When I go to boot the box, I get
> > "uncompressing kernel" and that's it.
> >
> > Thoughts?
>
> I'm getting at out of memory here; haven't looked why yet .. catching up
> again with linux-kernel to make sure it's not a known problem :)
Reply to: