![[LWN Logo]](/images/lwn.banner.gif)
Subject: BadRAM patch for 2.2.15
To: linux-kernel@vger.rutgers.edu (Linux Kernel)
Date: Thu, 18 May 2000 15:51:49 +0200 (CEST)
From: "Rick van Rein" <vanrein@zonnet.nl>
Linus,
Thanx for Linux!
This is my addition --- BadRAM, a patch to handle broken memory modules.
Many people enjoy it, and the cost (in terms of runtime performance) is 0.
For more information, refer to http://home.zonnet.nl/vanrein/badram
Appended, you will find version 3.1 of the patch, which is the stable V3
applied to kernel 2.2.15. Please be so kind to include it in future 2.2
kernels.
Truly yours,
Rick van Rein.
------------ 8< ------------ 8< ------------ 8< ------------ 8< ------------
diff -u -r -N linux-2.2.15.orig/CREDITS linux/CREDITS
--- linux-2.2.15.orig/CREDITS Tue Jan 4 19:12:10 2000
+++ linux/CREDITS Mon Mar 20 23:10:13 2000
@@ -1672,6 +1672,14 @@
S: Malvern, Pennsylvania 19355
S: USA
+N: Rick van Rein
+E: vanrein@cs.utwente.nl
+W: http://www.cs.utwente.nl/~vanrein
+D: Memory, the BadRAM subsystem dealing with statically challanged RAM modules.
+S: Binnenes 67
+S: 9407 CX Assen
+S: The Netherlands
+
N: Stefan Reinauer
E: stepan@linux.de
W: http://www.freiburg.linux.de/~stepan/
diff -u -r -N linux-2.2.15.orig/Documentation/Configure.help linux/Documentation/Configure.help
--- linux-2.2.15.orig/Documentation/Configure.help Tue Jan 4 19:12:10 2000
+++ linux/Documentation/Configure.help Sat Apr 8 01:19:32 2000
@@ -12145,6 +12145,21 @@
If you do not have any Quicknet telephony cards, you can safely
ignore this option.
+
+Work around bad spots in RAM
+CONFIG_BADRAM
+ This small kernel extension makes it possible to use memory chips
+ which are not entirely correct. It works by never allocating the
+ places that are wrong. Those places are specified with the badram
+ boot option to LILO. Read /usr/src/linux/Documentation/badram.txt
+ and/or visit http://home.zonnet.nl/vanrein/badram for information.
+
+ This option co-operates well with a second boot option from LILO
+ that starts memtest86, which is able to automatically produce the
+ patterns for the commandline in case of memory trouble.
+
+ It is safe to say 'Y' here, and it is advised because there is no
+ performance impact.
CPiA Video For Linux
CONFIG_VIDEO_CPIA
@@ -12399,5 +12414,5 @@
# LocalWords: BVME BVME WRITETHROUGH copyback writethrough fwmark syncookie tu
# LocalWords: alphalinux GOBIOS csn chemnitz nat ACARD AMI MegaRAID megaraid
# LocalWords: QNX4FS ISI isicom xterms Apollos VPN RCPCI rcpci sgi visws pcmcia
-# LocalWords: IrLPT UIRCC Tecra
+# LocalWords: IrLPT UIRCC Tecra BadRAM badram vanrein zonnet
diff -u -r -N linux-2.2.15.orig/Documentation/badram.txt linux/Documentation/badram.txt
--- linux-2.2.15.orig/Documentation/badram.txt Thu Jan 1 01:00:00 1970
+++ linux/Documentation/badram.txt Sat Apr 8 01:32:14 2000
@@ -0,0 +1,212 @@
+INFORMATION ON USING BAD RAM MODULES
+====================================
+
+Introduction
+ RAM is getting smaller and smaller, and as a result, also more and more
+ vulnerable. This makes the manufacturing of hardware more expensive,
+ since an excessive amount of RAM chips must be discarded on account of
+ a single cell that is wrong. Similarly, static discharge may damage a
+ RAM module forever, which is usually remedied by replacing it
+ entirely.
+
+ This is not necessary, as the BadRAM code shows: By informing the Linux
+ kernel which addresses in a RAM are damaged, the kernel simply avoids
+ ever allocating such addresses but makes all the rest available.
+
+Reasons for this feature
+ There are many reasons why this kernel feature is useful:
+ - Chip manufacture is resource intensive; waste less and sleep better
+ - It's another chance to promote Linux as "the flexible OS"
+ - Some laptops have their RAM soldered in... and then it fails!
+ - It's plain cool ;-)
+
+Running example
+ To run this project, I was given two DIMMs, 32 MB each. One, that we
+ shall use as a running example in this text, contained 512 faulty bits,
+ spread over 1/4 of the address range in a regular pattern. Some tricks
+ with a RAM tester and a few binary calculations were sufficient to
+ write these faults down in 2 longword numbers.
+
+ The kernel recognised the correct number of pages with faults and did
+ not give them out for allocation. The allocation routines could
+ therefore progress as normally, without any adaption.
+ So, I gained 30 MB of DIMM which would otherwise have been thrown
+ away. After booting the kernel, the kernel behaved exactly as it
+ always had.
+
+Initial checks
+ If you experience RAM trouble, first read /usr/src/linux/memory.txt
+ and try out the mem=4M trick to see if at least some initial parts
+ of your RAM work well. The BadRAM routines halt the kernel in panic
+ if the reserved area of memory (containing kernel stuff) contains
+ a faulty address.
+
+Running a RAM checker
+ The memory checker is not built into the kernel, to avoid delays at
+ runtime. If you experience problems that may be caused by RAM, run
+ a good RAM checker, such as
+ http://reality.sgi.com/cbrady_denver/memtest86
+ The output of a RAM checker provides addresses that went wrong. In
+ the 32 MB chip with 512 faulty bits mentioned above, the errors were
+ found in the 8MB-16MB range (the DIMM was in slot #0) at addresses
+ xxx42f4
+ xxx62f4
+ xxxc2f4
+ xxxe2f4
+ and the error was a "sticky 1 bit", a memory bit that stayed "1" no
+ matter what was written to it. The regularity of this pattern
+ suggests the death of a buffer at the output stages of a row on one of
+ the chips. I expect such regularity to be commonplace. Finding this
+ regularity currently is human effort, but it should not be hard to
+ alter a RAM checker to capture it in some sort of pattern, possibly
+ the BadRAM patterns described below.
+
+ By the way, if you manage to get hold of memtest86 version 2.3 or
+ beyond, you can configure the printing mode to produce BadRAM patterns,
+ which find out exactly what you must enter on the LILO: commandline,
+ except that you shouldn't mention the added spacing. That means that
+ you can skip the following step, which saves you a *lot* of work.
+
+Capturing errors in a pattern
+ Instead of manually providing all 512 errors to the kernel, it's nicer
+ to generate a pattern. Since the regularity is based on address decoding
+ software, which generally takes certain bits into account and ignores
+ others, we shall provide a faulty address F, together with a bit mask M
+ that specifies which bits must be equal to F. In C code, an address A
+ is faulty if and only if
+ (F & M) == (A & M)
+ or alternately (closer to a hardware implementation):
+ ~((F ^ A) & M)
+ In the example 32 MB chip, we had the faulty addresses in 8MB-16MB:
+ xxx42f4 ....0100....
+ xxx62f4 ....0110....
+ xxxc2f4 ....1100....
+ xxxe2f4 ....1110....
+ The second column represents the alternating hex digit in binary form.
+ Apperantly, the first and one-but last binary digit can be anything,
+ so the binary mask for that part is 0101. The mask for the part after
+ this is 0xfff, and the part before should select anything in the range
+ 8MB-16MB, or 0x00800000-0x01000000; this is done with a bitmask
+ 0xff80xxxx. Combining these partial masks, we get:
+ F=0x008042f4 M=0xff805fff
+ That covers everything for this DIMM; for more complicated failing
+ DIMMs, or for a combination of multiple failing DIMMs, it can be
+ necessary to set up a number of such F/M pairs.
+
+Rebooting Linux
+ Now that these patterns are known (and double-checked, the calculations
+ are highly error-prone... it would be neat to test them in the RAM
+ checker...) we simply restart Linux with these F/M pairs as a parameter.
+ If you normally boot as follows:
+ LILO: linux
+ you should now boot with
+ LILO: linux badram=0x008042f4,0xff805fff
+ or perhaps by mentioning more F/M pairs in an order F0,M0,F1,M1,...
+ When you provide an odd number of arguments to badram, the default mask
+ 0xffffffff (only one address matched) is applied to the pattern.
+
+ Beware of the commandline length. At least up to LILO version 0.21,
+ the commandline is cut off after the 78th character; later versions
+ may go as far as the kernel goes, namely 255 characters. In no way is
+ it possible to enter more than 10 numbers to the badram boot option.
+
+ When the kernel now boots, it should not give any trouble with RAM.
+ Mind you, this is under the assumption that the kernel and its data
+ storage do not overlap an erroneous part. If this happens, and the
+ kernel does not choke on it right away, it will stop with a panic.
+ You will need to provide a RAM where the initial, say 2MB, is faultless.
+
+ Now look up your memory status with
+ dmesg | grep ^Memory:
+ which prints a line much like
+ Memory: 158524k/163840k available
+ (940k kernel code,
+ 412k reserved,
+ 1856k data,
+ 60k init,
+ 2048k badram)
+ The latter entry, the badram, is 2048k to represent the loss of 2MB
+ of general purpose RAM due to the errors. Or, positively rephrased,
+ instead of throwing out 32MB as useless, you only throw out 2MB.
+
+ If the system is stable (try compiling a few kernels, and do a few
+ finds in / or so) you may add the boot parameter to /etc/lilo.conf
+ as a line to _all_ the kernels that handle this trouble with a line
+ append="badram=0x008042f4,0xff805fff"
+ after which you run "lilo".
+ Warning: Don't experiment with these settings on your only boot image.
+ If the BadRAM overlays kernel code, data, init, or other reserved
+ memory, the kernel will halt in panic. Try settings on a test boot
+ image first, and if you get a panic you should change the order of
+ your DIMMs [which may involve buying a new one just to be able to
+ change the order].
+
+BadRAM classification
+ This technique may start a lively market for "dead" RAM. It is important
+ to realise that some RAMs are more dead than others. So, instead of
+ just providing a RAM size, it is also important to know the BadRAM
+ class, which is defined as follows:
+
+ A BadRAM class N means that at most 2^N bytes have a problem,
+ and that all problems with the RAMs are persistent: They
+ are predictable and always show up.
+
+ The DIMM that serves as an example here was of class 9, since 512=2^9
+ errors were found. Higher classes are worse, "correct" RAM is of class
+ -1 (or even less, at your choice).
+ Class N also means that the bitmask for your chip (if there's just one,
+ that is) counts N bits "0" and it means that (if no faults fall in the
+ same page) an amount of 2^N*PAGESIZE memory is lost, in the example on
+ an i386 architecture that would be 2^9*4k=2MB, which accounts for the
+ initial claim of 30MB RAM gained with this DIMM.
+
+Known Bugs
+ LILO is known to cut off commandlines which are too long. For the
+ lilo-0.21 distribution, a commandline may not exceed 78 characters,
+ while actually, 255 would be possible [on i386, kernel 2.2.15].
+ LILO does _not_ report too-long commandlines, but the error will
+ show up as either a panic at boot time, stating
+ panic: BadRAM page in initial area
+ or the dmesg line starting with Memory: will mention an unpredicted
+ number of kilobytes. (Note that the latter number only includes
+ errors in accessed memory.)
+
+Future Possibilities
+ It would be possible to use even more of the faulty RAMs by employing
+ them for slabs. The smaller allocation granularity of slabs makes it
+ possible to throw out just, say, 32 bytes surrounding an error. This
+ would mean that the example DIMM only looses 16kB instead of 2MB.
+ It might even be possible to allocate the slabs in such a way that,
+ where possible, the remaining bytes in a slab structure are allocated
+ around the error, reducing the RAM loss to 0 in the optimal situation!
+
+ However, this yield is somewhat faked: It is possible to provide 512
+ pages of 32-byte slabs, but it is not certain that anyone would use
+ that many 32-byte slabs at any time.
+
+ A better solution might be to alter the page allocation for a slab to
+ have a preference for BadRAM pages, and given those a special treatment.
+ This way, the BadRAM would be spread over all the slabs, which seems
+ more likely to be a `true' pay-off. This would yield more overhead at
+ slab allocation time, but on the other hand, by the nature of slabs,
+ such allocations are made as rare as possible, so it might not matter
+ that much. I am uncertain where to go.
+
+Names and Places
+ The home page of this project is on
+ http://home.zonnet.nl/vanrein/badram
+ In case you have experiences with the BadRAM software which differ from
+ the test reportings on that site, I hope you will mail me with that
+ new information.
+
+ The BadRAM project is an idea and implementation by
+ Rick van Rein
+ Binnenes 67
+ 9407 CX Assen
+ The Netherlands
+ vanrein@cs.utwente.nl
+ If you like it, a postcard would be much appreciated ;-)
+
+
+ Enjoy,
+ -Rick.
diff -u -r -N linux-2.2.15.orig/Documentation/kernel-parameters.txt linux/Documentation/kernel-parameters.txt
--- linux-2.2.15.orig/Documentation/kernel-parameters.txt Tue Jan 4 19:12:10 2000
+++ linux/Documentation/kernel-parameters.txt Sat Apr 8 00:48:00 2000
@@ -13,6 +13,7 @@
APIC APIC support is enabled.
APM Advanced Power Management support is enabled.
AX25 Appropriate AX.25 support is enabled.
+ BADRAM Support for faulty RAM chips is enabled.
CD Appropriate CD support is enabled.
EIDE EIDE/ATAPI support is enabled.
FB The frame buffer device is enabled.
@@ -75,6 +76,8 @@
atascsi= [HW,SCSI] Atari SCSI.
aztcd= [HW,CD] Aztec CD driver.
+
+ badram= [BADRAM] Avoid allocating faulty RAM addresses.
baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem.
diff -u -r -N linux-2.2.15.orig/Documentation/memory.txt linux/Documentation/memory.txt
--- linux-2.2.15.orig/Documentation/memory.txt Wed Oct 27 02:53:39 1999
+++ linux/Documentation/memory.txt Mon Mar 20 23:33:00 2000
@@ -18,6 +18,13 @@
as you add more memory. Consider exchanging your
motherboard.
+ 4) A static discharge or production fault causes a RAM module
+ to have (predictable) errors, usually meaning that certain
+ bits cannot be set or reset. Instead of throwing away your
+ RAM module, you may read /usr/src/linux/Documentation/badram.txt
+ to learn how to detect, locate and circuimvent such errors
+ in your RAM module.
+
All of these problems can be addressed with the "mem=XXXM" boot option
(where XXX is the size of RAM to use in megabytes).
It can also tell Linux to use less memory than is actually installed.
@@ -45,6 +52,8 @@
* Try passing the "mem=4M" option to the kernel to limit
Linux to using a very small amount of memory.
+ If this helps, read /usr/src/linux/Documentation/badram.txt
+ to learn how to find and circuimvent memory errors.
Other tricks:
diff -u -r -N linux-2.2.15.orig/arch/i386/config.in linux/arch/i386/config.in
--- linux-2.2.15.orig/arch/i386/config.in Tue Jan 4 19:12:11 2000
+++ linux/arch/i386/config.in Sat Apr 8 01:18:05 2000
@@ -116,6 +116,8 @@
bool ' Use real mode APM BIOS call to power off' CONFIG_APM_REAL_MODE_POWER_OFF
fi
+bool 'Work around bad spots in RAM' CONFIG_BADRAM
+
endmenu
source drivers/pnp/Config.in
diff -u -r -N linux-2.2.15.orig/arch/i386/defconfig linux/arch/i386/defconfig
--- linux-2.2.15.orig/arch/i386/defconfig Thu Aug 26 02:29:46 1999
+++ linux/arch/i386/defconfig Sat Apr 8 01:04:38 2000
@@ -58,6 +58,7 @@
CONFIG_BINFMT_MISC=y
# CONFIG_PARPORT is not set
# CONFIG_APM is not set
+CONFIG_BADRAM=y
#
# Plug and Play support
diff -u -r -N linux-2.2.15.orig/arch/i386/mm/init.c linux/arch/i386/mm/init.c
--- linux-2.2.15.orig/arch/i386/mm/init.c Wed Oct 27 02:53:39 1999
+++ linux/arch/i386/mm/init.c Sat Apr 8 01:25:35 2000
@@ -393,6 +393,7 @@
int reservedpages = 0;
int datapages = 0;
int initpages = 0;
+ int badpages = 0;
unsigned long tmp;
end_mem &= PAGE_MASK;
@@ -426,10 +427,20 @@
clear_bit(PG_reserved, &mem_map[MAP_NR(start_mem)].flags);
start_mem += PAGE_SIZE;
}
+
+#ifdef CONFIG_BADRAM
+ badram_markpages ();
+#endif
+
for (tmp = PAGE_OFFSET ; tmp < end_mem ; tmp += PAGE_SIZE) {
if (tmp >= MAX_DMA_ADDRESS)
clear_bit(PG_DMA, &mem_map[MAP_NR(tmp)].flags);
if (PageReserved(mem_map+MAP_NR(tmp))) {
+#ifdef CONFIG_BADRAM
+ if (PageBad(mem_map+MAP_NR(tmp))) {
+ panic ("BadRAM page in initial area");
+ }
+#endif
if (tmp >= (unsigned long) &_text && tmp < (unsigned long) &_edata) {
if (tmp < (unsigned long) &_etext)
codepages++;
@@ -445,6 +456,12 @@
reservedpages++;
continue;
}
+#ifdef CONFIG_BADRAM
+ if (PageBad(mem_map+MAP_NR(tmp))) {
+ badpages++;
+ continue;
+ }
+#endif
atomic_set(&mem_map[MAP_NR(tmp)].count, 1);
#ifdef CONFIG_BLK_DEV_INITRD
if (!initrd_start || (tmp < initrd_start || tmp >=
@@ -452,13 +469,14 @@
#endif
free_page(tmp);
}
- printk("Memory: %luk/%luk available (%dk kernel code, %dk reserved, %dk data, %dk init)\n",
+ printk("Memory: %luk/%luk available (%dk kernel code, %dk reserved, %dk data, %dk init, %dk badram)\n",
(unsigned long) nr_free_pages << (PAGE_SHIFT-10),
max_mapnr << (PAGE_SHIFT-10),
codepages << (PAGE_SHIFT-10),
reservedpages << (PAGE_SHIFT-10),
datapages << (PAGE_SHIFT-10),
- initpages << (PAGE_SHIFT-10));
+ initpages << (PAGE_SHIFT-10),
+ badpages << (PAGE_SHIFT-10));
if (boot_cpu_data.wp_works_ok < 0)
test_wp_bit();
diff -u -r -N linux-2.2.15.orig/include/linux/mm.h linux/include/linux/mm.h
--- linux-2.2.15.orig/include/linux/mm.h Tue Jan 4 19:12:25 2000
+++ linux/include/linux/mm.h Fri Apr 7 23:27:21 2000
@@ -13,6 +13,12 @@
extern void * high_memory;
extern int page_cluster;
+// The maximum number of values (address, mask pairs) covering bad RAM.
+// This value is hardcoded (!) in init/main.c, get_options()
+#define BADRAM_MAXPARAMS 10
+extern void badram_setup (char *str, int *ints);
+void badram_markpages (void);
+
#include <asm/page.h>
#include <asm/atomic.h>
@@ -144,6 +150,7 @@
#define PG_Slab 9
#define PG_swap_cache 10
#define PG_skip 11
+#define PG_badram 13
#define PG_reserved 31
/* Make it prettier to test the above... */
@@ -159,6 +166,7 @@
#define PageSlab(page) (test_bit(PG_Slab, &(page)->flags))
#define PageSwapCache(page) (test_bit(PG_swap_cache, &(page)->flags))
#define PageReserved(page) (test_bit(PG_reserved, &(page)->flags))
+#define PageBad(page) (test_bit(PG_badram, &(page)->flags))
#define PageSetSlab(page) (set_bit(PG_Slab, &(page)->flags))
#define PageSetSwapCache(page) (set_bit(PG_swap_cache, &(page)->flags))
diff -u -r -N linux-2.2.15.orig/init/main.c linux/init/main.c
--- linux-2.2.15.orig/init/main.c Tue Jan 4 19:12:25 2000
+++ linux/init/main.c Sat Apr 8 01:26:49 2000
@@ -22,6 +22,7 @@
#include <linux/smp_lock.h>
#include <linux/blk.h>
#include <linux/hdreg.h>
+#include <linux/mm.h>
#include <asm/io.h>
#include <asm/bugs.h>
@@ -983,6 +984,9 @@
#endif
#ifdef CONFIG_BLK_CPQ_DA
{ "smart2=", cpqarray_setup },
+#endif
+#ifdef CONFIG_BADRAM
+ { "badram=", badram_setup },
#endif
{ 0, 0 }
};
diff -u -r -N linux-2.2.15.orig/mm/page_alloc.c linux/mm/page_alloc.c
--- linux-2.2.15.orig/mm/page_alloc.c Tue Jan 4 19:12:26 2000
+++ linux/mm/page_alloc.c Sat Apr 8 01:27:52 2000
@@ -331,6 +331,72 @@
return start_mem;
}
+
+#ifdef CONFIG_BADRAM
+
+
+/* Given a pointed-at address and a mask, increment the page so that the
+ * mask hides the increment. Return 0 if no increment is possible.
+ */
+static int next_masked_address (ulong *addrp, ulong mask)
+{
+ ulong inc=1;
+ ulong newval = *addrp;
+ while (inc & mask)
+ inc += inc;
+ while (inc != 0) {
+ newval += inc;
+ newval &= ~mask;
+ newval |= ((*addrp) & mask);
+ if (newval > *addrp) {
+ *addrp = newval;
+ return 1;
+ }
+ do {
+ inc += inc;
+ } while (inc & ~mask);
+ while (inc & mask)
+ inc += inc;
+ }
+ return 0;
+}
+
+
+static ulong badram_params [BADRAM_MAXPARAMS+1];
+
+void badram_setup (char *str, int *argv)
+{
+ int i;
+ if (argv[0] > BADRAM_MAXPARAMS)
+ panic ("Too bad: badram=... length exceeds BADRAM_MAXPARAMS. "
+ "Try to combine patterns");
+ for (i=0; i<=argv[0]; i++) {
+ badram_params [i]=argv [i];
+ }
+}
+
+void badram_markpages (void) {
+ ulong *argv=badram_params;
+ int argc=*argv++;
+ ulong addr, mask;
+ while (argc-- > 0) {
+ addr = *argv++;
+ mask = (argc-- > 0) ? *argv++ : ~0L;
+ mask |= ~PAGE_MASK; // Optimalisation
+ addr &= mask; // Normalisation
+ do {
+ if ( ( addr >> PAGE_SHIFT ) >= max_mapnr)
+ break;
+ set_bit (PG_badram,
+ &mem_map [ addr >> PAGE_SHIFT ].flags);
+ } while (next_masked_address (&addr,mask));
+ }
+}
+
+
+#endif /* CONFIG_BADRAM */
+
+
/*
* Primitive swap readahead code. We simply read an aligned block of
* (1 << page_cluster) entries in the swap area. This method is chosen
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/