From mboxrd@z Thu Jan 1 00:00:00 1970 Received: with ECARTIS (v1.0.0; list linux-mips); Thu, 09 Dec 2004 22:49:55 +0000 (GMT) Received: from xes-inc.com ([IPv6:::ffff:24.196.136.110]:51330 "EHLO xes-inc.com") by linux-mips.org with ESMTP id ; Thu, 9 Dec 2004 22:49:48 +0000 Received: from matts ([10.52.0.13]) by xes-inc.com (8.11.6/8.11.6) with SMTP id iB9MnjN11321; Thu, 9 Dec 2004 16:49:45 -0600 Message-ID: <062301c4de41$5bf43cb0$0d00340a@matts> From: "Matthew Starzewski" To: Cc: Subject: Using more than 256 MB of memory on SB1250 in 32-bit mode, revisited Date: Thu, 9 Dec 2004 16:49:35 -0600 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0620_01C4DE0F.10321590" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 6.00.2720.3000 X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2600.0000 Return-Path: X-Envelope-To: <"|/home/ecartis/ecartis -s linux-mips"> (uid 0) X-Orcpt: rfc822;linux-mips@linux-mips.org Original-Recipient: rfc822;linux-mips@linux-mips.org X-archive-position: 6627 X-ecartis-version: Ecartis v1.0.0 Sender: linux-mips-bounce@linux-mips.org Errors-to: linux-mips-bounce@linux-mips.org X-original-sender: mstarzewski@xes-inc.com Precedence: bulk X-list: linux-mips This is a multi-part message in MIME format. ------=_NextPart_000_0620_01C4DE0F.10321590 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I've tried to enable HIGHMEM to access all 512MB of SDRAM on a BCM1125 based board as per this previous thread: Using more than 256 MB of memory on SB1250 in 32-bit mode : http://www.spinics.net/lists/mips/msg14396.html BCM1125 Board: XPedite3000 PrPMC http://www.xes-inc.com/Products/XPedite/XPedite3000/XPedite3000.html In MIPS32 mode, the memory map comes out to the following: Physical Memory Map: 0x00000000 - 0x0FFFFFFF 256MB 0x80000000 - 0x8FFFFFFF 256MB Virtual Memory Map: 0x80000000 - 0x8FFFFFFF 256MB <<<< INACCESSIBLE >>>> 256MB My goal in enabling HIGHMEM was to get at the upper 256MB much as described here. Access to the upper 256MB through HIGHMEM may incur a performance hit, but it's certainly better going without. http://home.earthlink.net/~jknapka/linux-mm/vminit.html#PAGING_INIT However, what I get is a stall as per the log pasted below. Enabling CONFIG_64BIT_PHYS_ADDR does not make a difference. Results were cross-verified between CFE and another bootloader. When I hand the physical memory described above off to = add_memory_region, I noticed a few odd things. One thing that looks suspicious is that the = variable void *high_memory ends up being set to 0xa0000000, or right at KSEG1, in mm/init.c. Also, num_physpages became huge, 0x90000, because the init code in kernel/setup.c and mm/init.c want a page for every memory location, even highmem. Is this appropriate when the memory is not directly accessible via __va and virt_to_phys? =20 This may be an ancillary effect of what's mentioned above, but when = num_physpages grows in size, nr_free_pages doesn't track with it, so in void = vfs_caches_init(unsigned long mempages) and the like, you get a horrible underflow condition: /* code */ printk("MJS - nr_free_pages():0x%X\n", nr_free_pages()); printk("MJS - OLD mempages:0x%X\n", mempages); reserve =3D (mempages - nr_free_pages()) * 3/2; mempages -=3D reserve; printk("MJS - NEW reserve:0x%X mempages:0x%X\n", reserve, mempages); /* printout */ MJS - nr_free_pages():0x1E9A0 MJS - OLD mempages:0x90000 MJS - NEW reserve:0xAA190 mempages:0xFFFE5E70 Let me know what you think of this issue. In anticipation of the "Why not use MIPS64 build?" question, we'd prefer to and will, but the=20 MIPS64 build has underperformed or had bugs (SATA seek time, networking signals, etc) that MIPS32 doesn't. For these issues or any case where the MIPS64 build isn't there yet, it makes sense to have the MIPS32 path open. Regards, Matt =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D Kernel Boot Log = =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D CFE version 1.0.37 for BCM91125E (64bit,SP,BE) Build Date: Mon Jun 28 18:33:29 CDT 2004 (mstarzewski@lsys1) Copyright (C) 2000,2001,2002,2003 Broadcom Corporation. Initializing Arena. Initializing Devices. BCM91125E board revision 1 Config switch: 0 CPU: 1125H A2 L2 Cache: 256KB SysCfg: 00800000004A0600 [PLL_DIV: 12, IOB0_DIV: CPUCLK/4, IOB1_DIV: = CPUCLK/3] CPU type 0x40103: 600MHz Total memory: 0x20000000 bytes (512MB) Total memory used by CFE: 0x8FE8AFC0 - 0x90000000 (1527872) Initialized Data: 0x8FE8AFC0 - 0x8FE94D40 (40320) BSS Area: 0x8FE94D40 - 0x8FE95430 (1776) Local Heap: 0x8FE95430 - 0x8FF95430 (1048576) Stack Area: 0x8FF95430 - 0x8FF97430 (8192) Text (code) segment: 0x8FF97440 - 0x8FFFFFB8 (428920) Boot area (physical): 0x0FE49000 - 0x0FE89000 Relocation Factor: I:F0397440 - D:0DF8AFC0 CFE> ifconfig eth0 -addr=3D10.52.33.67 -mask=3D255.255.0.0;boot -elf = 10.52.0.4:/home/mstarzewski/li nux/kernels/mips26_works_cfe/vmlinux eth0: Link speed: 100BaseT FDX Device eth0: hwaddr 40-00-10-06-40-00, ipaddr 10.52.33.67, mask = 255.255.0.0 gateway not set, nameserver not set Loader:elf Filesys:tftp Dev:eth0 = File:10.52.0.4:/home/mstarzewski/linux/kernels/mips26_works_cf e/vmlinux Options:(null) Loading: 0xffffffff80100000/2558312 0xffffffff80370968/146032 Entry at = 0xffffffff8032f018 Variable Name Value -------------------- -------------------------------------------------- BOOT_CONSOLE uart0 CPU_TYPE 1125H CPU_REVISION A2 CPU_NUM_CORES 1 CPU_SPEED 600 CFE_VERSION 1.0.37 CFE_BOARDNAME BCM91125E CFE_MEMORYSIZE 512 NET_DEVICE eth0 NET_IPADDR 10.52.33.67 NET_NETMASK 255.255.0.0 NET_GATEWAY 0.0.0.0 NET_NAMESERVER 0.0.0.0 BOOT_DEVICE eth0 BOOT_FILE = 10.52.0.4:/home/mstarzewski/linux/kernels/mips26_works_cfe/vmlinux DELETING BOOT_CONSOLE DELETING CPU_REVISION DELETING CPU_SPEED DELETING CFE_BOARDNAME DELETING NET_DEVICE DELETING NET_NETMASK DELETING NET_NAMESERVER DELETING BOOT_FILE Variable Name Value -------------------- -------------------------------------------------- Closing network. Starting program at 0xffffffff8032f018 CFE addr:0x000000000FE8A000 size:0x0000000000000000, type:0 CFE addr:0x0000000010000000 size:0x0000000000000000, type:-2147483648 Broadcom SiByte BCM1125H A2 @ 600 MHz (SB1 rev 3) Board type: SiByte BCM91250A (SWARM) Linux version 2.6.6-rc3 (mstarzewski@lsys1) (gcc version 3.2.3 with = SiByte modifications) #17 S MP Thu Dec 9 11:08:23 CST 2004 CPU revision is: 00040103 FPU revision is: 000f0103 This kernel optimized for board runs with CFE Determined physical RAM map: memory: 0fe89e00 @ 00000000 (usable) memory: 0ffffe00 @ 80000000 (usable) 1791MB HIGHMEM available. On node 0 totalpages: 589823 DMA zone: 131072 pages, LIFO batch:16 Normal zone: 0 pages, LIFO batch:1 HighMem zone: 458751 pages, LIFO batch:16 Built 1 zonelists Kernel command line: root=3D/dev/nfs ip=3Dauto rw PID hash table entries: 4096 (order 12: 32768 bytes) Memory: 495040k/260644k available (1850k kernel code, 27020k reserved, = 382k data, 264k init, 26 2140k highmem) Calibrating delay loop... 399.36 BogoMIPS << STALL >> ------=_NextPart_000_0620_01C4DE0F.10321590 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
I've tried to enable HIGHMEM to access = all 512MB=20 of
SDRAM on a BCM1125 based board as per this previous
thread:
 
Using more than=20 256 MB of memory on SB1250 in 32-bit mode :
http://www.spini= cs.net/lists/mips/msg14396.html
BCM1125 Board: XPedite3000 PrPMC
http://www.xes-inc.com/Products/XPedite/XPedite3000/XPedite3000.html=
 
In MIPS32 mode, the memory map comes = out to the=20 following:
 
Physical Memory Map:
0x00000000 - 0x0FFFFFFF  =20 256MB
0x80000000 - 0x8FFFFFFF  =20 256MB
Virtual Memory Map:
0x80000000 - 0x8FFFFFFF  =20 256MB
<<<< INACCESSIBLE = >>>>=20 256MB
 
My goal in enabling HIGHMEM was to get = at the upper=20 256MB
much as described here.  Access to = the upper=20 256MB through
HIGHMEM may incur a performance hit, = but it's=20 certainly better
going without.
 
http://home.earthlink.net/~jknapka/linux-mm/vminit.html#PAGING_INIT
 
However, what I get is a stall as per = the log=20 pasted below.  Enabling
CONFIG_64BIT_PHYS_ADDR does not make a=20 difference.  Results
were cross-verified between CFE and = another=20 bootloader.
 
When I hand the=20 physical memory described above off to add_memory_region,
I noticed a few odd things. One thing = that looks=20 suspicious is that the variable
void = *high_memory ends up=20 being set to 0xa0000000, or right at KSEG1,
in mm/init.c.
 
Also, num_physpages became huge, = 0x90000, because=20 the init code in
kernel/setup.c and mm/init.c want a = page for every=20 memory location, even
highmem.  Is this appropriate when = the memory=20 is not directly accessible
via __va and virt_to_phys?  =
 
This may be an ancillary effect of = what's mentioned=20 above, but when num_physpages
grows in size, nr_free_pages doesn't = track with it,=20 so in void vfs_caches_init(unsigned long
mempages) and the like, you get a = horrible=20 underflow condition:
 
/* code */
       =20 printk("MJS - nr_free_pages():0x%X\n",=20 nr_free_pages());
        = printk("MJS -=20 OLD mempages:0x%X\n", mempages);
      =  =20 reserve =3D (mempages - nr_free_pages()) *=20 3/2;
        mempages -=3D=20 reserve;
        printk("MJS - NEW = reserve:0x%X=20 mempages:0x%X\n",
        &nbs= p;     =20 reserve, mempages);

/* printout */
MJS - nr_free_pages():0x1E9A0
MJS - = OLD=20 mempages:0x90000
MJS - NEW reserve:0xAA190=20 mempages:0xFFFE5E70
 
Let me know what you think of this = issue.  In=20 anticipation of the
"Why not use MIPS64 build?" question, = we'd prefer=20 to and will, but the
MIPS64 build has = underperformed or had=20 bugs (SATA seek time,
networking signals, etc) = that MIPS32=20 doesn't.  For = these=20 issues
or any case where the MIPS64 build = isn't there yet,=20 it makes
sense to have the MIPS32 path = open.
 
Regards,
Matt
 
 
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D = Kernel Boot Log=20 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
 
CFE version 1.0.37 for BCM91125E=20 (64bit,SP,BE)
Build Date: Mon Jun 28 18:33:29 CDT 2004 (mstarzewski@lsys1)
Copyright (C)=20 2000,2001,2002,2003 Broadcom Corporation.
 
Initializing Arena.
Initializing=20 Devices.
BCM91125E board revision 1
Config switch: 0
CPU: 1125H = A2
L2 Cache: 256KB
SysCfg: 00800000004A0600 [PLL_DIV: 12, = IOB0_DIV:=20 CPUCLK/4, IOB1_DIV: CPUCLK/3]
CPU type 0x40103: 600MHz
Total = memory:=20 0x20000000 bytes (512MB)
 
Total memory used by CFE:  = 0x8FE8AFC0 -=20 0x90000000 (1527872)
Initialized=20 Data:          0x8FE8AFC0 - = 0x8FE94D40 (40320)
BSS=20 Area:           &n= bsp;     =20 0x8FE94D40 - 0x8FE95430 (1776)
Local=20 Heap:           &n= bsp;   =20 0x8FE95430 - 0x8FF95430 (1048576)
Stack=20 Area:           &n= bsp;   =20 0x8FF95430 - 0x8FF97430 (8192)
Text (code)=20 segment:       0x8FF97440 - 0x8FFFFFB8=20 (428920)
Boot area (physical):      = 0x0FE49000 -=20 0x0FE89000
Relocation = Factor:        =20 I:F0397440 - D:0DF8AFC0
 
CFE> ifconfig eth0 = -addr=3D10.52.33.67=20 -mask=3D255.255.0.0;boot -elf=20 10.52.0.4:/home/mstarzewski/li
nux/kernels/mips26_works_cfe/vmlinuxeth0:=20 Link speed: 100BaseT FDX
Device eth0:  hwaddr 40-00-10-06-40-00, = ipaddr=20 10.52.33.67, mask = 255.255.0.0
       =20 gateway not set, nameserver not set
Loader:elf Filesys:tftp Dev:eth0=20 File:10.52.0.4:/home/mstarzewski/linux/kernels/mips26_works_cf
e/vmlin= ux=20 Options:(null)
Loading: 0xffffffff80100000/2558312 = 0xffffffff80370968/146032=20 Entry at 0xffffffff8032f018
Variable=20 Name        = Value
--------------------=20 --------------------------------------------------
BOOT_CONSOLE &= nbsp;      =20 uart0
CPU_TYPE         &n= bsp;  =20 1125H
CPU_REVISION        =20 A2
CPU_NUM_CORES       =20 1
CPU_SPEED          = ; =20 600
CFE_VERSION          = 1.0.37
CFE_BOARDNAME       =20 BCM91125E
CFE_MEMORYSIZE      =20 512
NET_DEVICE         &n= bsp;=20 eth0
NET_IPADDR         &= nbsp;=20 10.52.33.67
NET_NETMASK        = ; =20 255.255.0.0
NET_GATEWAY        = ; =20 0.0.0.0
NET_NAMESERVER      =20 0.0.0.0
BOOT_DEVICE        &nb= sp;=20 eth0
BOOT_FILE         &n= bsp; =20 10.52.0.4:/home/mstarzewski/linux/kernels/mips26_works_cfe/vmlinux
DEL= ETING=20 BOOT_CONSOLE
DELETING CPU_REVISION
DELETING CPU_SPEED
DELETING=20 CFE_BOARDNAME
DELETING NET_DEVICE
DELETING NET_NETMASK
DELETING = NET_NAMESERVER
DELETING BOOT_FILE
Variable=20 Name        = Value
--------------------=20 --------------------------------------------------
Closing=20 network.
Starting program at 0xffffffff8032f018
CFE=20 addr:0x000000000FE8A000 size:0x0000000000000000, type:0
CFE=20 addr:0x0000000010000000 size:0x0000000000000000, = type:-2147483648
Broadcom=20 SiByte BCM1125H A2 @ 600 MHz (SB1 rev 3)
Board type: SiByte BCM91250A = (SWARM)
Linux version 2.6.6-rc3 (mstarzewski@lsys1) (gcc = version 3.2.3=20 with SiByte modifications) #17 S
MP Thu Dec 9 11:08:23 CST = 2004
CPU=20 revision is: 00040103
FPU revision is: 000f0103
This kernel = optimized for=20 board runs with CFE
Determined physical RAM map:
 memory: = 0fe89e00 @=20 00000000 (usable)
 memory: 0ffffe00 @ 80000000 = (usable)
1791MB=20 HIGHMEM available.
On node 0 totalpages: 589823
  DMA zone: = 131072=20 pages, LIFO batch:16
  Normal zone: 0 pages, LIFO = batch:1
 =20 HighMem zone: 458751 pages, LIFO batch:16
Built 1 zonelists
Kernel = command=20 line: root=3D/dev/nfs ip=3Dauto rw
PID hash table entries: 4096 = (order 12: 32768=20 bytes)
Memory: 495040k/260644k available (1850k kernel code, 27020k = reserved,=20 382k data, 264k init, 26
2140k highmem)
Calibrating delay loop... = 399.36=20 BogoMIPS
 
<< STALL=20 >>
------=_NextPart_000_0620_01C4DE0F.10321590-- From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <062301c4de41$5bf43cb0$0d00340a@matts> From: "Matthew Starzewski" Subject: Using more than 256 MB of memory on SB1250 in 32-bit mode, revisited Date: Thu, 9 Dec 2004 16:49:35 -0600 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0620_01C4DE0F.10321590" Return-Path: Sender: linux-mips-bounce@linux-mips.org Errors-to: linux-mips-bounce@linux-mips.org To: linux-mips@linux-mips.org Cc: Steve.Finney@SpirentCom.COM Message-ID: <20041209224935.baSUtuHDFtewFqi-ROeq9w85rcCanrb9FdxCpuBs6qE@z> This is a multi-part message in MIME format. ------=_NextPart_000_0620_01C4DE0F.10321590 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I've tried to enable HIGHMEM to access all 512MB of SDRAM on a BCM1125 based board as per this previous thread: Using more than 256 MB of memory on SB1250 in 32-bit mode : http://www.spinics.net/lists/mips/msg14396.html BCM1125 Board: XPedite3000 PrPMC http://www.xes-inc.com/Products/XPedite/XPedite3000/XPedite3000.html In MIPS32 mode, the memory map comes out to the following: Physical Memory Map: 0x00000000 - 0x0FFFFFFF 256MB 0x80000000 - 0x8FFFFFFF 256MB Virtual Memory Map: 0x80000000 - 0x8FFFFFFF 256MB <<<< INACCESSIBLE >>>> 256MB My goal in enabling HIGHMEM was to get at the upper 256MB much as described here. Access to the upper 256MB through HIGHMEM may incur a performance hit, but it's certainly better going without. http://home.earthlink.net/~jknapka/linux-mm/vminit.html#PAGING_INIT However, what I get is a stall as per the log pasted below. Enabling CONFIG_64BIT_PHYS_ADDR does not make a difference. Results were cross-verified between CFE and another bootloader. When I hand the physical memory described above off to = add_memory_region, I noticed a few odd things. One thing that looks suspicious is that the = variable void *high_memory ends up being set to 0xa0000000, or right at KSEG1, in mm/init.c. Also, num_physpages became huge, 0x90000, because the init code in kernel/setup.c and mm/init.c want a page for every memory location, even highmem. Is this appropriate when the memory is not directly accessible via __va and virt_to_phys? =20 This may be an ancillary effect of what's mentioned above, but when = num_physpages grows in size, nr_free_pages doesn't track with it, so in void = vfs_caches_init(unsigned long mempages) and the like, you get a horrible underflow condition: /* code */ printk("MJS - nr_free_pages():0x%X\n", nr_free_pages()); printk("MJS - OLD mempages:0x%X\n", mempages); reserve =3D (mempages - nr_free_pages()) * 3/2; mempages -=3D reserve; printk("MJS - NEW reserve:0x%X mempages:0x%X\n", reserve, mempages); /* printout */ MJS - nr_free_pages():0x1E9A0 MJS - OLD mempages:0x90000 MJS - NEW reserve:0xAA190 mempages:0xFFFE5E70 Let me know what you think of this issue. In anticipation of the "Why not use MIPS64 build?" question, we'd prefer to and will, but the=20 MIPS64 build has underperformed or had bugs (SATA seek time, networking signals, etc) that MIPS32 doesn't. For these issues or any case where the MIPS64 build isn't there yet, it makes sense to have the MIPS32 path open. Regards, Matt =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D Kernel Boot Log = =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D CFE version 1.0.37 for BCM91125E (64bit,SP,BE) Build Date: Mon Jun 28 18:33:29 CDT 2004 (mstarzewski@lsys1) Copyright (C) 2000,2001,2002,2003 Broadcom Corporation. Initializing Arena. Initializing Devices. BCM91125E board revision 1 Config switch: 0 CPU: 1125H A2 L2 Cache: 256KB SysCfg: 00800000004A0600 [PLL_DIV: 12, IOB0_DIV: CPUCLK/4, IOB1_DIV: = CPUCLK/3] CPU type 0x40103: 600MHz Total memory: 0x20000000 bytes (512MB) Total memory used by CFE: 0x8FE8AFC0 - 0x90000000 (1527872) Initialized Data: 0x8FE8AFC0 - 0x8FE94D40 (40320) BSS Area: 0x8FE94D40 - 0x8FE95430 (1776) Local Heap: 0x8FE95430 - 0x8FF95430 (1048576) Stack Area: 0x8FF95430 - 0x8FF97430 (8192) Text (code) segment: 0x8FF97440 - 0x8FFFFFB8 (428920) Boot area (physical): 0x0FE49000 - 0x0FE89000 Relocation Factor: I:F0397440 - D:0DF8AFC0 CFE> ifconfig eth0 -addr=3D10.52.33.67 -mask=3D255.255.0.0;boot -elf = 10.52.0.4:/home/mstarzewski/li nux/kernels/mips26_works_cfe/vmlinux eth0: Link speed: 100BaseT FDX Device eth0: hwaddr 40-00-10-06-40-00, ipaddr 10.52.33.67, mask = 255.255.0.0 gateway not set, nameserver not set Loader:elf Filesys:tftp Dev:eth0 = File:10.52.0.4:/home/mstarzewski/linux/kernels/mips26_works_cf e/vmlinux Options:(null) Loading: 0xffffffff80100000/2558312 0xffffffff80370968/146032 Entry at = 0xffffffff8032f018 Variable Name Value -------------------- -------------------------------------------------- BOOT_CONSOLE uart0 CPU_TYPE 1125H CPU_REVISION A2 CPU_NUM_CORES 1 CPU_SPEED 600 CFE_VERSION 1.0.37 CFE_BOARDNAME BCM91125E CFE_MEMORYSIZE 512 NET_DEVICE eth0 NET_IPADDR 10.52.33.67 NET_NETMASK 255.255.0.0 NET_GATEWAY 0.0.0.0 NET_NAMESERVER 0.0.0.0 BOOT_DEVICE eth0 BOOT_FILE = 10.52.0.4:/home/mstarzewski/linux/kernels/mips26_works_cfe/vmlinux DELETING BOOT_CONSOLE DELETING CPU_REVISION DELETING CPU_SPEED DELETING CFE_BOARDNAME DELETING NET_DEVICE DELETING NET_NETMASK DELETING NET_NAMESERVER DELETING BOOT_FILE Variable Name Value -------------------- -------------------------------------------------- Closing network. Starting program at 0xffffffff8032f018 CFE addr:0x000000000FE8A000 size:0x0000000000000000, type:0 CFE addr:0x0000000010000000 size:0x0000000000000000, type:-2147483648 Broadcom SiByte BCM1125H A2 @ 600 MHz (SB1 rev 3) Board type: SiByte BCM91250A (SWARM) Linux version 2.6.6-rc3 (mstarzewski@lsys1) (gcc version 3.2.3 with = SiByte modifications) #17 S MP Thu Dec 9 11:08:23 CST 2004 CPU revision is: 00040103 FPU revision is: 000f0103 This kernel optimized for board runs with CFE Determined physical RAM map: memory: 0fe89e00 @ 00000000 (usable) memory: 0ffffe00 @ 80000000 (usable) 1791MB HIGHMEM available. On node 0 totalpages: 589823 DMA zone: 131072 pages, LIFO batch:16 Normal zone: 0 pages, LIFO batch:1 HighMem zone: 458751 pages, LIFO batch:16 Built 1 zonelists Kernel command line: root=3D/dev/nfs ip=3Dauto rw PID hash table entries: 4096 (order 12: 32768 bytes) Memory: 495040k/260644k available (1850k kernel code, 27020k reserved, = 382k data, 264k init, 26 2140k highmem) Calibrating delay loop... 399.36 BogoMIPS << STALL >> ------=_NextPart_000_0620_01C4DE0F.10321590 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
I've tried to enable HIGHMEM to access = all 512MB=20 of
SDRAM on a BCM1125 based board as per this previous
thread:
 
Using more than=20 256 MB of memory on SB1250 in 32-bit mode :
http://www.spini= cs.net/lists/mips/msg14396.html
BCM1125 Board: XPedite3000 PrPMC
http://www.xes-inc.com/Products/XPedite/XPedite3000/XPedite3000.html=
 
In MIPS32 mode, the memory map comes = out to the=20 following:
 
Physical Memory Map:
0x00000000 - 0x0FFFFFFF  =20 256MB
0x80000000 - 0x8FFFFFFF  =20 256MB
Virtual Memory Map:
0x80000000 - 0x8FFFFFFF  =20 256MB
<<<< INACCESSIBLE = >>>>=20 256MB
 
My goal in enabling HIGHMEM was to get = at the upper=20 256MB
much as described here.  Access to = the upper=20 256MB through
HIGHMEM may incur a performance hit, = but it's=20 certainly better
going without.
 
http://home.earthlink.net/~jknapka/linux-mm/vminit.html#PAGING_INIT
 
However, what I get is a stall as per = the log=20 pasted below.  Enabling
CONFIG_64BIT_PHYS_ADDR does not make a=20 difference.  Results
were cross-verified between CFE and = another=20 bootloader.
 
When I hand the=20 physical memory described above off to add_memory_region,
I noticed a few odd things. One thing = that looks=20 suspicious is that the variable
void = *high_memory ends up=20 being set to 0xa0000000, or right at KSEG1,
in mm/init.c.
 
Also, num_physpages became huge, = 0x90000, because=20 the init code in
kernel/setup.c and mm/init.c want a = page for every=20 memory location, even
highmem.  Is this appropriate when = the memory=20 is not directly accessible
via __va and virt_to_phys?  =
 
This may be an ancillary effect of = what's mentioned=20 above, but when num_physpages
grows in size, nr_free_pages doesn't = track with it,=20 so in void vfs_caches_init(unsigned long
mempages) and the like, you get a = horrible=20 underflow condition:
 
/* code */
       =20 printk("MJS - nr_free_pages():0x%X\n",=20 nr_free_pages());
        = printk("MJS -=20 OLD mempages:0x%X\n", mempages);
      =  =20 reserve =3D (mempages - nr_free_pages()) *=20 3/2;
        mempages -=3D=20 reserve;
        printk("MJS - NEW = reserve:0x%X=20 mempages:0x%X\n",
        &nbs= p;     =20 reserve, mempages);

/* printout */
MJS - nr_free_pages():0x1E9A0
MJS - = OLD=20 mempages:0x90000
MJS - NEW reserve:0xAA190=20 mempages:0xFFFE5E70
 
Let me know what you think of this = issue.  In=20 anticipation of the
"Why not use MIPS64 build?" question, = we'd prefer=20 to and will, but the
MIPS64 build has = underperformed or had=20 bugs (SATA seek time,
networking signals, etc) = that MIPS32=20 doesn't.  For = these=20 issues
or any case where the MIPS64 build = isn't there yet,=20 it makes
sense to have the MIPS32 path = open.
 
Regards,
Matt
 
 
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D = Kernel Boot Log=20 =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
 
CFE version 1.0.37 for BCM91125E=20 (64bit,SP,BE)
Build Date: Mon Jun 28 18:33:29 CDT 2004 (mstarzewski@lsys1)
Copyright (C)=20 2000,2001,2002,2003 Broadcom Corporation.
 
Initializing Arena.
Initializing=20 Devices.
BCM91125E board revision 1
Config switch: 0
CPU: 1125H = A2
L2 Cache: 256KB
SysCfg: 00800000004A0600 [PLL_DIV: 12, = IOB0_DIV:=20 CPUCLK/4, IOB1_DIV: CPUCLK/3]
CPU type 0x40103: 600MHz
Total = memory:=20 0x20000000 bytes (512MB)
 
Total memory used by CFE:  = 0x8FE8AFC0 -=20 0x90000000 (1527872)
Initialized=20 Data:          0x8FE8AFC0 - = 0x8FE94D40 (40320)
BSS=20 Area:           &n= bsp;     =20 0x8FE94D40 - 0x8FE95430 (1776)
Local=20 Heap:           &n= bsp;   =20 0x8FE95430 - 0x8FF95430 (1048576)
Stack=20 Area:           &n= bsp;   =20 0x8FF95430 - 0x8FF97430 (8192)
Text (code)=20 segment:       0x8FF97440 - 0x8FFFFFB8=20 (428920)
Boot area (physical):      = 0x0FE49000 -=20 0x0FE89000
Relocation = Factor:        =20 I:F0397440 - D:0DF8AFC0
 
CFE> ifconfig eth0 = -addr=3D10.52.33.67=20 -mask=3D255.255.0.0;boot -elf=20 10.52.0.4:/home/mstarzewski/li
nux/kernels/mips26_works_cfe/vmlinuxeth0:=20 Link speed: 100BaseT FDX
Device eth0:  hwaddr 40-00-10-06-40-00, = ipaddr=20 10.52.33.67, mask = 255.255.0.0
       =20 gateway not set, nameserver not set
Loader:elf Filesys:tftp Dev:eth0=20 File:10.52.0.4:/home/mstarzewski/linux/kernels/mips26_works_cf
e/vmlin= ux=20 Options:(null)
Loading: 0xffffffff80100000/2558312 = 0xffffffff80370968/146032=20 Entry at 0xffffffff8032f018
Variable=20 Name        = Value
--------------------=20 --------------------------------------------------
BOOT_CONSOLE &= nbsp;      =20 uart0
CPU_TYPE         &n= bsp;  =20 1125H
CPU_REVISION        =20 A2
CPU_NUM_CORES       =20 1
CPU_SPEED          = ; =20 600
CFE_VERSION          = 1.0.37
CFE_BOARDNAME       =20 BCM91125E
CFE_MEMORYSIZE      =20 512
NET_DEVICE         &n= bsp;=20 eth0
NET_IPADDR         &= nbsp;=20 10.52.33.67
NET_NETMASK        = ; =20 255.255.0.0
NET_GATEWAY        = ; =20 0.0.0.0
NET_NAMESERVER      =20 0.0.0.0
BOOT_DEVICE        &nb= sp;=20 eth0
BOOT_FILE         &n= bsp; =20 10.52.0.4:/home/mstarzewski/linux/kernels/mips26_works_cfe/vmlinux
DEL= ETING=20 BOOT_CONSOLE
DELETING CPU_REVISION
DELETING CPU_SPEED
DELETING=20 CFE_BOARDNAME
DELETING NET_DEVICE
DELETING NET_NETMASK
DELETING = NET_NAMESERVER
DELETING BOOT_FILE
Variable=20 Name        = Value
--------------------=20 --------------------------------------------------
Closing=20 network.
Starting program at 0xffffffff8032f018
CFE=20 addr:0x000000000FE8A000 size:0x0000000000000000, type:0
CFE=20 addr:0x0000000010000000 size:0x0000000000000000, = type:-2147483648
Broadcom=20 SiByte BCM1125H A2 @ 600 MHz (SB1 rev 3)
Board type: SiByte BCM91250A = (SWARM)
Linux version 2.6.6-rc3 (mstarzewski@lsys1) (gcc = version 3.2.3=20 with SiByte modifications) #17 S
MP Thu Dec 9 11:08:23 CST = 2004
CPU=20 revision is: 00040103
FPU revision is: 000f0103
This kernel = optimized for=20 board runs with CFE
Determined physical RAM map:
 memory: = 0fe89e00 @=20 00000000 (usable)
 memory: 0ffffe00 @ 80000000 = (usable)
1791MB=20 HIGHMEM available.
On node 0 totalpages: 589823
  DMA zone: = 131072=20 pages, LIFO batch:16
  Normal zone: 0 pages, LIFO = batch:1
 =20 HighMem zone: 458751 pages, LIFO batch:16
Built 1 zonelists
Kernel = command=20 line: root=3D/dev/nfs ip=3Dauto rw
PID hash table entries: 4096 = (order 12: 32768=20 bytes)
Memory: 495040k/260644k available (1850k kernel code, 27020k = reserved,=20 382k data, 264k init, 26
2140k highmem)
Calibrating delay loop... = 399.36=20 BogoMIPS
 
<< STALL=20 >>
------=_NextPart_000_0620_01C4DE0F.10321590--