Add extended crashkernel command line syntax
Subsystems (8)
cover Add extended crashkernel command line syntax
This patch adds a extended crashkernel syntax that makes the value of reserved system RAM dependent on the system RAM itself:
crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset]
range=start-[end]For example:
crashkernel=512M-2G:64M,2G-:128M
The motivation comes from distributors that configure their crashkernel command line automatically with some configuration tool (YaST, you know ;)). Of course that tool knows the value of System RAM, but if the user removes RAM, then the system becomes unbootable or at least unusable and error handling is very difficult.
This series implements this change for i386, x86_64, ia64, ppc64 and sh. That should be all platforms that support kdump in current mainline. I tested all platforms except sh due to the lack of a sh processor.
The patch series is against 2.6.23-rc8-mm1 and replaces following patches:
- extended-crashkernel-command-line.patch
- use-extended-crashkernel-command-line-on-i386.patch
- use-extended-crashkernel-command-line-on-i386-fix-config_nohighmem-for-extended-crashkernel-command-line.patch
- use-extended-crashkernel-command-line-on-i386-fix-config_nohighmem-for-extended-crashkernel-command-line-fix.patch
- use-extended-crashkernel-command-line-on-x86_64.patch
- use-extended-crashkernel-command-line-on-ia64.patch
- use-extended-crashkernel-command-line-on-ia64-fix.patch
- use-extended-crashkernel-command-line-on-ppc64.patch
- use-extended-crashkernel-command-line-on-sh.patch
- add-documentation-for-extended-crashkernel-syntax.patch
- add-documentation-for-extended-crashkernel-syntax-add-extended-crashkernel-syntax-to-kernel-parameterstxt.patch
Modifications compared to last submit:
- typecast to (unsigned long long) before shifting
- small coding style adjustments
- BUG_ON() in parse_crashkernel()
- merge patch that adds documentation do Documentation/kernel-parameters.txt in documentation patch (this has been an extra patch previously)
- fix build on IA64
- fix build on i386 with CONFIG_NOHIGHMEM
PATCH 1/7 Extended crashkernel command line
This is the generic part of the patch. It adds a parse_crashkernel() function in kernel/kexec.c that is called by the architecture specific code that actually reserves the memory. That function takes the whole command line and looks itself for "crashkernel=" in it.
If there are multiple occurrences, then the last one is taken. The advantage is that if you have a bootloader like lilo or elilo which allows you to append a command line parameter but not to remove one (like in GRUB), then you can add another crashkernel value for testing at the boot command line and this one overwrites the command line in the configuration then.
--- include/linux/kexec.h | 2 kernel/kexec.c | 142 ++++++++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 144 insertions(+)
include/linux/kexec.h +2 -0
@@ -187,6 +187,8 @@ extern u32 vmcoreinfo_note[VMCOREINFO_NOextern size_t vmcoreinfo_size;extern size_t vmcoreinfo_max_size; +int __init parse_crashkernel (char *cmdline,unsigned long long system_ram, +unsigned long long *crash_size,unsigned long long *crash_base);#else /* !CONFIG_KEXEC */ struct pt_regs ;
kernel/kexec.c +142 -0
148 lines not on this page · raw message
Re: [patch 1/7] Extended crashkernel command line
* Tue, 25 Sep 2007 20:22:58 +0200
Quoted 5 lines
This is the generic part of the patch. It adds a
parse_crashkernel()function inkernel/kexec.cthat is called by the architecture specific code that actually reserves the memory. That function takes the whole command line and looks itself for "crashkernel=" in it.
[]
include/linux/kexec.h extern u32 vmcoreinfo_note[VMCOREINFO_NO · line 191
Quoted 7 lines
+++
b/include/linux/kexec.hinclude/linux/kexec.h +2
-0@@ -187,6 +187,8 @@ extern u32 vmcoreinfo_note[VMCOREINFO_NOextern size_t vmcoreinfo_size;extern size_t vmcoreinfo_max_size; +int __init parse_crashkernel (char *cmdline,unsigned long long system_ram, +unsigned long long *crash_size,unsigned long long *crash_base);
I still want to point out my consernes about `unsigned long long long' bloat.
kernel/kexec.c
Quoted 4 lines
include/linux/kexec.h +0
-0#else /* !CONFIG_KEXEC */ struct pt_regs ;kernel/kexec.c +0
-0
[]
Quoted 4 lines
+/*
+ * This function parses command lines in the format
+ *
+ * crashkernel=<ramsize-range>:<size>[,...][@<base>]
#v+
olecom@flower:/tmp$ grep '@offset' <extended-crashkernel-cmdline.mbox
crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset]
- /* crashkernel=size@offset specifies the size to reserve for a crash
+While the "crashkernel=size[@offset]" syntax is sufficient for most
+ crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset]
+ crashkernel=range1:size1[,range2:size2,...][@offset]
olecom@flower:/tmp$ grep '@base' <extended-crashkernel-cmdline.mbox
+ * crashkernel=size[@base]
olecom@flower:/tmp$ grep '@<base' <extended-crashkernel-cmdline.mbox
+ * crashkernel=<ramsize-range>:<size>[,...][@<base>]
olecom@flower:/tmp$ grep '@<offset' <extended-crashkernel-cmdline.mbox
olecom@flower:/tmp$
#v-Quoted 4 lines
+ *
+ * The function returns 0 on success and-EINVALon failure.
+ */kernel/kexec.c +1
-0+static int __init parse_crashkernel_mem(char *cmdline,
kernel/kexec.c static int __init crash_notes_memory_ini · line 1174
Quoted 10 lines
kernel/kexec.c +10
-0+unsigned long long system_ram, + unsigned long long *crash_size, + unsigned long long *crash_base) +{ + char *cur = cmdline; + +/* for each entry of the comma-separated list */ + do { + unsigned long long start = 0, end = ULLONG_MAX; + unsigned long long size = -1;
memparse(), as a wrapper for somple_strtoll(), always have a return value (zero by default).
<http://article.gmane.org/gmane.linux.kernel/583426>
Consider coded error path now, please.
And why not to make overall result reliable? This is kernel after all.
I.e. if there's valid `crashkernel=' option, but some parsing errors, why not to apply some heuristics with warning in syslog, if user have some conf, bootloader (random) errors, but feature still works?
Quoted 36 lines
kernel/kexec.c +36
-0+ +/* get the start of the range */ + start =memparse (cur, &cur); +if (*cur !='-' ) { +printk (KERN_WARNING "crashkernel: '-' expected\n" ); +return -EINVAL ; + } + cur++; + +/* if no ':' is here, than we read the end */ +if (*cur !=':' ) { + end =memparse (cur, &cur); +if (end <= start) { +printk (KERN_WARNING "crashkernel: end <= start\n" ); +return -EINVAL ; + } + } + +if (*cur !=':' ) { +printk (KERN_WARNING "crashkernel: ':' expected\n" ); +return -EINVAL ; + } + cur++; + + size =memparse (cur, &cur); +if (size <0 ) { +printk (KERN_WARNING "crashkernel: invalid size\n" ); +return -EINVAL ; + } + +/* match ? */ +if (system_ram >= start && system_ram <= end) { + *crash_size = size; +break ; + } + }while (*cur++ ==',' );
kernel/kexec.c static int __init crash_notes_memory_ini · line 1220
Quoted 10 lines
kernel/kexec.c +10
-0+ +if (*crash_size >0 ) { +while (*cur !=' ' && *cur !='@' ) + cur++; +if (*cur =='@' ) + *crash_base =memparse (cur+1 , &cur); + } + +return 0 ; +}
--
-o--=O`C
#oo'L O
<___=E M
Re: [patch 1/7] Extended crashkernel command line
Quoted 7 lines · Oleg Verych
* Tue, 25 Sep 2007 20:22:58 +0200
Quoted 6 lines
+++
b/include/linux/kexec.h(quoted diff) +0
-0@@ -187,6 +187,8 @@ extern u32 vmcoreinfo_note[VMCOREINFO_NO extern size_t vmcoreinfo_size; extern size_t vmcoreinfo_max_size;
Quoted 5 lines
Review extern u32 vmcoreinfo_note[VMCOREINFO_NO · line 191
Quoted 2 lines
(quoted diff) +2
-0+int __init parse_crashkernel(char *cmdline, unsigned long long system_ram, + unsigned long long *crash_size, unsigned long long *crash_base);I still want to point out my consernes about `unsigned long long long' bloat.
What "concerns" (it's unsigned long long and not unsigned long long long)? Is is common coding style in the Linux kernel to *not* use unsigned long long? This type *is* used e.g. in arch/i386/kernel/e820.c also for pysical memory values.
Quoted 7 lines
#v+
olecom@flower:/tmp$ grep '@offset' <extended-crashkernel-cmdline.mbox crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset](quoted diff) +3
-1- /* crashkernel=size@offset specifies the size to reserve for a crash +While the "crashkernel=size[@offset]" syntax is sufficient for most + crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] + crashkernel=range1:size1[,range2:size2,...][@offset]
Quoted 7 lines
olecom@flower:/tmp$ grep '@base' <extended-crashkernel-cmdline.mbox + * crashkernel=size[@base] olecom@flower:/tmp$ grep '@<base' <extended-crashkernel-cmdline.mbox + * crashkernel=<ramsize-range>:<size>[,...][@<base>] olecom@flower:/tmp$ grep '@<offset' <extended-crashkernel-cmdline.mbox olecom@flower:/tmp$ #v-
The patch below fixes this.
Quoted 10 lines
Quoted 10 lines
(quoted diff) +10
-0+ * + * The function returns 0 on success and -EINVAL on failure. + */ +static int __init parse_crashkernel_mem(char *cmdline, + unsigned long long system_ram, + unsigned long long *crash_size, + unsigned long long *crash_base) +{ + char *cur = cmdline; +
Quoted 8 lines
Review
Quoted 4 lines
(quoted diff) +4
-0+ /* for each entry of the comma-separated list */ + do { + unsigned long long start = 0, end = ULLONG_MAX; + unsigned long long size = -1;
memparse(), as a wrapper forsomple_strtoll(), always have a return value (zero by default).
<http://article.gmane.org/gmane.linux.kernel/583426>
Next reply (because of a different patch).
---
Only use 'offset' and not 'base' for the address where the reserved area starts.
--- kernel/kexec.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
kernel/kexec.c +2 -2
@@ -1157,7 +1157,7 @@ module_init(crash_notes_memory_init)/* * This function parses command lines in the format * -* crashkernel=<ramsize-range>:<size>[,...][@<base>] +* crashkernel=ramsize-range:size[,...][@offset] * * The function returns 0 on success and -EINVAL on failure. */
@@ -1222,7 +1222,7 @@ static int __init parse_crashkernel_mem(/* * That function parses "simple" (old) crashkernel command lines like * -* crashkernel=size[@base] +* crashkernel=size[@offset] * * It returns 0 on success and -EINVAL on failure. */
Re: [patch 1/7] Extended crashkernel command line
In reply to Oleg VerychQuoted 11 lines · Oleg Verych
Quoted 11 lines
+ *
+ * The function returns 0 on success and-EINVALon failure.
+ */(quoted diff) +7
-0+static int __init parse_crashkernel_mem(char *cmdline, + unsigned long long system_ram, + unsigned long long *crash_size, + unsigned long long *crash_base) +{ + char *cur = cmdline; +
Quoted 8 lines
Review
Quoted 4 lines
(quoted diff) +4
-0+ /* for each entry of the comma-separated list */ + do { + unsigned long long start = 0, end = ULLONG_MAX; + unsigned long long size = -1;
memparse(), as a wrapper forsomple_strtoll(), always have a return value (zero by default).
<http://article.gmane.org/gmane.linux.kernel/583426>
Ok, that's fixed now, see patch below.
Quoted 5 lines
And why not to make overall result reliable? This is kernel after all.
I.e. if there's valid `crashkernel=' option, but some parsing errors, why not to apply some heuristics with warning in syslog, if user have some conf, bootloader (random) errors, but feature still works?
I'm against guessing. The user has to specify a parameter which is right according to syntax.
However, I plan to make a patch that the kernel can detect a sensible offset automatically for i386 and x86_64 as it's done in ia64. Since both architectures have a relocatable kernel now, that makes perfectly sense. But that's another patch.
---
This patches improves error handling in parse_crashkernel_mem() by comparing the return pointer of memparse() with the input pointer and also replaces all printk(KERN_WARNING msg) with pr_warning(msg).
--- kernel/kexec.c | 31 ++++++++++++++++++++++++------- 1 file changed, 24 insertions(+), 7 deletions(-)
kernel/kexec.c +24 -7
@@ -1172,33 +1172,50 @@ static int __init parse_crashkernel_mem( do { unsigned long long start = 0, end = ULLONG_MAX; unsigned long long size = -1; + char *tmp; /* get the start of the range */ - start = memparse(cur, &cur); + start = memparse(cur, &tmp); + if (cur == tmp) { + pr_warning("crashkernel: Memory value expected\n"); + return -EINVAL; + } + cur = tmp; if (*cur != '-') { - printk(KERN_WARNING "crashkernel: '-' expected\n"); + pr_warning("crashkernel: '-' expected\n"); return -EINVAL; } cur++; /* if no ':' is here, than we read the end */ if (*cur != ':') { - end = memparse(cur, &cur); + end = memparse(cur, &tmp); + if (cur == tmp) { + pr_warning("crashkernel: Memory " + "value expected\n"); + return -EINVAL; + } + cur = tmp; if (end <= start) { - printk(KERN_WARNING "crashkernel: end <= start\n"); + pr_warning("crashkernel: end <= start\n"); return -EINVAL; } } if (*cur != ':') { - printk(KERN_WARNING "crashkernel: ':' expected\n"); + pr_warning("crashkernel: ':' expected\n"); return -EINVAL; } cur++; - size = memparse(cur, &cur); + size = memparse(cur, &tmp); + if (cur == tmp) { + pr_warning("Memory value expected\n"); + return -EINVAL; + } + cur = tmp; if (size < 0) { - printk(KERN_WARNING "crashkernel: invalid size\n"); + pr_warning("crashkernel: invalid size\n"); return -EINVAL; }
Re: [patch 1/7] Extended crashkernel command line
Wed, Sep 26, 2007 at 06:16:02PM +0200, Bernhard Walle (part two, see bottom):
Quoted 3 lines
Quoted 3 lines
memparse(), as a wrapper forsomple_strtoll(), always have a return value (zero by default).
<http://article.gmane.org/gmane.linux.kernel/583426>
Sorry for my typos, i should write `simple_strtoull()'. This function (ULL from str) have always return value grater or equal to zero.
Thus,
kernel/kexec.c static int __init parse_crashkernel_mem( · line 1174
Quoted 13 lines
Signed-off-by Bernhard Walle <[email protected]>--- kernel/kexec.c | 31 ++++++++++++++++++++++++------- 1 file changed, 24 insertions(+), 7 deletions(-)kernel/kexec.c +0
-0@@ -1172,33 +1172,50 @@ static int __init parse_crashkernel_mem( do { unsigned long long start = 0, end = ULLONG_MAX; unsigned long long size = -1;
no need in assigning values here, unless you plan to use them in case of `return -EINVAL', but i can not see that,
Quoted 39 lines
kernel/kexec.c +17
-5+char *tmp;/* get the start of the range */ - start =memparse (cur, &cur); + start =memparse (cur, &tmp); +if (cur == tmp) { +pr_warning ("crashkernel: Memory value expected\n" ); +return -EINVAL ; + } + cur = tmp;if (*cur !='-' ) { -printk (KERN_WARNING "crashkernel: '-' expected\n" ); +pr_warning ("crashkernel: '-' expected\n" );return -EINVAL ; } cur++;/* if no ':' is here, than we read the end */ if (*cur !=':' ) { - end =memparse (cur, &cur); + end =memparse (cur, &tmp); +if (cur == tmp) { +pr_warning ("crashkernel: Memory " +"value expected\n" ); +return -EINVAL ; + } + cur = tmp;if (end <= start) { -printk (KERN_WARNING "crashkernel: end <= start\n" ); +pr_warning ("crashkernel: end <= start\n" );return -EINVAL ; } }if (*cur !=':' ) { -printk (KERN_WARNING "crashkernel: ':' expected\n" ); +pr_warning ("crashkernel: ':' expected\n" );return -EINVAL ; }
kernel/kexec.c static int __init parse_crashkernel_mem( · line 1217
Quoted 10 lines
kernel/kexec.c +6
-1cur++; - size =memparse (cur, &cur); + size =memparse (cur, &tmp); +if (cur == tmp) { +pr_warning ("Memory value expected\n" ); +return -EINVAL ; + } + cur = tmp;if (size <0 ) {
`size' cannot be less that zero here (wonder, if it matters to have userspace model of this parser and actually feed it with garbage input).
kernel/kexec.c static int __init parse_crashkernel_mem( · line 1221
Quoted 5 lines
kernel/kexec.c +1
-1-printk (KERN_WARNING "crashkernel: invalid size\n" ); +pr_warning ("crashkernel: invalid size\n" );return -EINVAL ; }
Wed, Sep 26, 2007 at 06:16:02PM +0200, Bernhard Walle (part one):
Quoted 8 lines
Ok, that's fixed now, see patch below.
Quoted 5 lines
And why not to make overall result reliable? This is kernel after all.
I.e. if there's valid `crashkernel=' option, but some parsing errors, why not to apply some heuristics with warning in syslog, if user have some conf, bootloader (random) errors, but feature still works?
Quoted 7 lines
I'm against guessing. The user has to specify a parameter which is right according to syntax.
However, I plan to make a patch that the kernel can detect a sensible offset automatically for i386 and
x86_64as it's done in ia64. Since both architectures have a relocatable kernel now, that makes perfectly sense. But that's another patch.
I was thinking about errors in YaST or typos in bootloader config, that may appear sometimes. And kernel must tolerate this kind of userspace input to be more reliable. But you know better, i just am waving hands.
____
("Mail-Followup-To:" respected)
Re: [patch 1/7] Extended crashkernel command line
Quoted 4 lines · Oleg Verych
I was thinking about errors in YaST or typos in bootloader config, that may appear sometimes. And kernel must tolerate this kind of userspace input to be more reliable. But you know better, i just am waving hands.
Of course the kernel must be able to handle them -- outputting an error message that can be read by the user and not crashing. But I don't expect that the kernel then reservate crash memory by guessing of values.
Thanks,
Bernhard
Re: [patch 1/7] Extended crashkernel command line
In reply to Oleg VerychQuoted 10 lines · Oleg Verych
kernel/kexec.c static int __init parse_crashkernel_mem( · line 1174
Quoted 7 lines
kernel/kexec.c +0
-0@@ -1172,33 +1172,50 @@ static int __init parse_crashkernel_mem( do { unsigned long long start = 0, end = ULLONG_MAX; unsigned long long size = -1;no need in assigning values here, unless you plan to use them in case of `return
-EINVAL', but i can not see that,
What about this (and yes, I tested with some wrong strings with Qemu):
----
This patch improves error handling in parse_crashkernel_mem() by comparing the return pointer of memparse() with the input pointer and also replaces all printk(KERN_WARNING msg) with pr_warning(msg).
--- kernel/kexec.c | 54 +++++++++++++++++++++++++++++++++++++++--------------- 1 file changed, 39 insertions(+), 15 deletions(-)
kernel/kexec.c +39 -15
@@ -1167,44 +1167,59 @@ static int __init parse_crashkernel_mem(unsigned long long *crash_size, unsigned long long *crash_base) { - char *cur = cmdline; + char *cur = cmdline, *tmp;/* for each entry of the comma-separated list */ do { - unsigned long long start = 0, end = ULLONG_MAX; - unsigned long long size = -1; + unsigned long long start, end = ULLONG_MAX, size; /* get the start of the range */ - start = memparse(cur, &cur); + start = memparse(cur, &tmp); + if (cur == tmp) { + pr_warning("crashkernel: Memory value expected\n"); + return -EINVAL; + } + cur = tmp; if (*cur != '-') { - printk(KERN_WARNING "crashkernel: '-' expected\n"); + pr_warning("crashkernel: '-' expected\n"); return -EINVAL; } cur++; /* if no ':' is here, than we read the end */ if (*cur != ':') { - end = memparse(cur, &cur); + end = memparse(cur, &tmp); + if (cur == tmp) { + pr_warning("crashkernel: Memory " + "value expected\n"); + return -EINVAL; + } + cur = tmp; if (end <= start) { - printk(KERN_WARNING "crashkernel: end <= start\n"); + pr_warning("crashkernel: end <= start\n"); return -EINVAL; } } if (*cur != ':') { - printk(KERN_WARNING "crashkernel: ':' expected\n"); + pr_warning("crashkernel: ':' expected\n"); return -EINVAL; } cur++; - size = memparse(cur, &cur); - if (size < 0) { - printk(KERN_WARNING "crashkernel: invalid size\n"); + size = memparse(cur, &tmp); + if (cur == tmp) { + pr_warning("Memory value expected\n"); + return -EINVAL; + } + cur = tmp; + if (size >= system_ram) { + pr_warning("crashkernel: invalid size\n"); return -EINVAL; } /* match ? */ - if (system_ram >= start && system_ram <= end) { + if (system_ram >= start && system_ram <= end) { *crash_size = size; break; }
@@ -1213,8 +1228,15 @@ static int __init parse_crashkernel_mem(if (*crash_size >0 ) {while (*cur !=' ' && *cur !='@' ) cur++; -if (*cur =='@' ) - *crash_base =memparse (cur+1 , &cur); +if (*cur =='@' ) { + cur++; + *crash_base =memparse (cur, &tmp); +if (cur == tmp) { +pr_warning ("Memory value expected " +"after '@'\n" ); +return -EINVAL ; + } + } }return 0 ;
@@ -1234,8 +1256,10 @@ static int __init parse_crashkernel_simpchar *cur = cmdline; *crash_size =memparse (cmdline, &cur); -if (cmdline == cur) +if (cmdline == cur) { +pr_warning ("crashkernel: memory value expected\n" );return -EINVAL ; + }if (*cur =='@' ) *crash_base =memparse (cur+1 , &cur);
Oleg Verych
Reviewed-byreviewed (Re: [patch 1/7] Extended crashkernel command line)
Wed, Sep 26, 2007 at 11:05:33PM +0200, Bernhard Walle:
Quoted 12 lines · Oleg Verych
Quoted 10 lines
kernel/kexec.c static int __init parse_crashkernel_mem( · line 1174
Quoted 7 lines
kernel/kexec.c +0
-0@@ -1172,33 +1172,50 @@ static int __init parse_crashkernel_mem( do { unsigned long long start = 0, end = ULLONG_MAX; unsigned long long size = -1;no need in assigning values here, unless you plan to use them in case of `return
-EINVAL', but i can not see that,
Quoted 1 line
What about this (and yes, I tested with some wrong strings with Qemu):
Thanks :D
Quoted 8 lines
----
This patch improves error handling in
parse_crashkernel_mem()by comparing the return pointer ofmemparse()with the input pointer and also replaces allprintk(KERN_WARNING msg)withpr_warning(msg).Signed-off-by Bernhard Walle <[email protected]>
_____
PATCH 2/7 Use extended crashkernel command line on i386
In reply to Bernhard WalleThis patch removes the crashkernel parsing from arch/i386/kernel/machine_kexec.c and calls the generic function, introduced in the last patch, in setup_bootmem_allocator().
This is necessary because the amount of System RAM must be known in this function now because of the new syntax.
--- arch/i386/kernel/e820.c | 3 +- arch/i386/kernel/machine_kexec.c | 22 ----------------- arch/i386/kernel/setup.c | 49 +++++++++++++++++++++++++++++++++++---- 3 files changed, 46 insertions(+), 28 deletions(-)
Files a/arch/i386/kernel/.setup.c.swp and b/arch/i386/kernel/.setup.c.swp differ
arch/i386/kernel/e820.c +2 -1
@@ -288,7 +288,8 @@ legacy_init_iomem_resources(struct resourequest_resource (res, code_resource);request_resource (res, data_resource);#ifdef CONFIG_KEXEC -request_resource (res, &crashk_res); +if (crashk_res.start != crashk_res.end) +request_resource (res, &crashk_res);#endif } }
arch/i386/kernel/machine_kexec.c +0 -22
@@ -149,28 +149,6 @@ NORET_TYPE void machine_kexec(struct kim image->start, cpu_has_pae); } -/* crashkernel=size@addr specifies the location to reserve for -* a crash kernel. By reserving this memory we guarantee -* that linux never sets it up as a DMA target. -* Useful for holding code to do something appropriate -* after a kernel panic. -*/ -static int __init parse_crashkernel (char *arg) -{ -unsigned long size, base; - size =memparse (arg, &arg); -if (*arg =='@' ) { - base =memparse (arg+1 , &arg); -/* FIXME: Do I want a sanity check -* to validate the memory range? -*/ - crashk_res.start = base; - crashk_res.end = base + size -1 ; - } -return 0 ; -} -early_param ("crashkernel" , parse_crashkernel); -void arch_crash_save_vmcoreinfo (void ) {#ifdef CONFIG_ARCH_DISCONTIGMEM_ENABLE
arch/i386/kernel/setup.c +44 -5
61 lines not on this page · raw message
PATCH 3/7 Use extended crashkernel command line on x86_64
In reply to Bernhard WalleThis patch removes the crashkernel parsing from arch/x86_64/kernel/machine_kexec.c and calls the generic function, introduced in the last patch, in setup_bootmem_allocator().
This is necessary because the amount of System RAM must be known in this function now because of the new syntax.
--- arch/x86_64/kernel/e820.c | 3 +- arch/x86_64/kernel/machine_kexec.c | 27 ------------------------- arch/x86_64/kernel/setup.c | 39 ++++++++++++++++++++++++++++++------- 3 files changed, 34 insertions(+), 35 deletions(-)
arch/x86_64/kernel/e820.c +2 -1
@@ -226,7 +226,8 @@ void __init e820_reserve_resources(void)request_resource (res, &code_resource);request_resource (res, &data_resource);#ifdef CONFIG_KEXEC -request_resource (res, &crashk_res); +if (crashk_res.start != crashk_res.end) +request_resource (res, &crashk_res);#endif } }
arch/x86_64/kernel/machine_kexec.c +0 -27
@@ -231,33 +231,6 @@ NORET_TYPE void machine_kexec(struct kim image->start); } -/* crashkernel=size@addr specifies the location to reserve for -* a crash kernel. By reserving this memory we guarantee -* that linux never set's it up as a DMA target. -* Useful for holding code to do something appropriate -* after a kernel panic. -*/ -static int __init setup_crashkernel (char *arg) -{ -unsigned long size, base; -char *p; -if (!arg) -return -EINVAL ; - size =memparse (arg, &p); -if (arg == p) -return -EINVAL ; -if (*p =='@' ) { - base =memparse (p+1 , &p); -/* FIXME: Do I want a sanity check to validate the -* memory range? Yes you do, but it's too early for -* e820 -AK */ - crashk_res.start = base; - crashk_res.end = base + size -1 ; - } -return 0 ; -} -early_param ("crashkernel" , setup_crashkernel); -void arch_crash_save_vmcoreinfo (void ) {#ifdef CONFIG_ARCH_DISCONTIGMEM_ENABLE
arch/x86_64/kernel/setup.c +32 -7
@@ -194,6 +194,37 @@ static inline void copy_edd(void) }#endif +#ifdef CONFIG_KEXEC +static void __init reserve_crashkernel (void ) +{ +unsigned long long free_mem; +unsigned long long crash_size, crash_base; +int ret; + + free_mem = ((unsigned long long )max_low_pfn - min_low_pfn) <<PAGE_SHIFT ; + + ret =parse_crashkernel (boot_command_line, free_mem, + &crash_size, &crash_base); +if (ret ==0 && crash_size) { +if (crash_base >0 ) { +printk (KERN_INFO "Reserving %ldMB of memory at %ldMB " +"for crashkernel (System RAM: %ldMB)\n" , + (unsigned long )(crash_size >>20 ), + (unsigned long )(crash_base >>20 ), + (unsigned long )(free_mem >>20 )); + crashk_res.start = crash_base; + crashk_res.end = crash_base + crash_size -1 ; +reserve_bootmem (crash_base, crash_size); + }else +printk (KERN_INFO "crashkernel reservation failed - " +"you have to specify a base address\n" ); + } +} +#else +static inline void __init reserve_crashkernel (void ) +{} +#endif +#define EBDA_ADDR_POINTER 0x40Eunsigned __initdata ebda_addr;
@@ -365,13 +396,7 @@ void __init setup_arch(char **cmdline_p) } }#endif -#ifdef CONFIG_KEXEC -if (crashk_res.start != crashk_res.end) { -reserve_bootmem_generic (crashk_res.start, - crashk_res.end - crashk_res.start +1 ); - } -#endif - +reserve_crashkernel ();paging_init ();#ifdef CONFIG_PCI
PATCH 4/7 Use extended crashkernel command line on ia64
In reply to Bernhard WalleThis patch adapts IA64 to use the generic parse_crashkernel() function instead of its own parsing for the crashkernel command line.
Because the total amount of System RAM must be known when calling this function, efi_memmap_init() is modified to return its accumulated total_memory variable.
Also, the crashkernel handling is moved in an own function in arch/ia64/kernel/setup.c to make the code more readable.
--- arch/ia64/kernel/efi.c | 4 +- arch/ia64/kernel/setup.c | 88 +++++++++++++++++++++++---------------------- include/asm-ia64/meminit.h | 2 - 3 files changed, 50 insertions(+), 44 deletions(-)
arch/ia64/kernel/efi.c +3 -1
@@ -967,7 +967,7 @@ find_memmap_space (void)* to use. We can allocate partial granules only if the unavailable * parts exist, and are WB. */ -void +unsigned long efi_memmap_init (unsigned long *s,unsigned long *e) {struct kern_memdesc *k, *prev =NULL ;
@@ -1084,6 +1084,8 @@ efi_memmap_init(unsigned long *s, unsign/* reserve the memory we are using for kern_memmap */ *s = (u64 )kern_memmap; *e = (u64 )++k; + +return total_memory; }void
arch/ia64/kernel/setup.c +46 -42
108 lines not on this page · raw message
include/asm-ia64/meminit.h +1 -1
@@ -35,7 +35,7 @@ extern void find_memory (void);extern void reserve_memory (void );extern void find_initrd (void );extern int filter_rsvd_memory (unsigned long start,unsigned long end,void *arg); -extern void efi_memmap_init (unsigned long *,unsigned long *); +extern unsigned long efi_memmap_init (unsigned long *s,unsigned long *e);extern int find_max_min_low_pfn (unsigned long ,unsigned long ,void *);extern unsigned long vmcore_find_descriptor_size (unsigned long address);
PATCH 5/7 Use extended crashkernel command line on ppc64
In reply to Bernhard WalleThis patch adapts the ppc64 code to use the generic parse_crashkernel() function introduced in the generic patch of that series.
--- arch/powerpc/kernel/machine_kexec.c | 52 ++++++++++++++++++------------------ 1 file changed, 26 insertions(+), 26 deletions(-)
arch/powerpc/kernel/machine_kexec.c +26 -26
@@ -61,45 +61,39 @@ NORET_TYPE void machine_kexec(struct kimfor (;;); } -static int __init early_parse_crashk (char *p) +void __init reserve_crashkernel (void ) { -unsigned long size; - -if (!p) -return 1 ; - - size =memparse (p, &p); +unsigned long long crash_size, crash_base; +int ret; -if (*p =='@' ) - crashk_res.start =memparse (p +1 , &p); -else - crashk_res.start =KDUMP_KERNELBASE ; - - crashk_res.end = crashk_res.start + size -1 ; - -return 0 ; -} -early_param ("crashkernel" , early_parse_crashk); +/* this is necessary because of lmb_phys_mem_size() */ +lmb_analyze (); -void __init reserve_crashkernel (void ) -{ -unsigned long size; +/* use common parsing */ + ret =parse_crashkernel (boot_command_line,lmb_phys_mem_size (), + &crash_size, &crash_base); +if (ret ==0 && crash_size >0 ) { +if (crash_base ==0 ) + crash_base =KDUMP_KERNELBASE ; + crashk_res.start = crash_base; + }else { +/* handle the device tree */ + crash_size = crashk_res.end - crashk_res.start +1 ; + } -if (crashk_res.start ==0 ) +if (crash_size ==0 )return ;/* We might have got these values via the command line or the * device tree, either way sanitise them now. */ - size = crashk_res.end - crashk_res.start +1 ; -if (crashk_res.start !=KDUMP_KERNELBASE )printk ("Crash kernel location must be 0x%x\n" ,KDUMP_KERNELBASE ); crashk_res.start =KDUMP_KERNELBASE ; - size =PAGE_ALIGN (size); - crashk_res.end = crashk_res.start + size -1 ; + crash_size =PAGE_ALIGN (crash_size); + crashk_res.end = crashk_res.start + crash_size -1 ;/* Crash kernel trumps memory limit */ if (memory_limit && memory_limit <= crashk_res.end) {
@@ -108,7 +102,13 @@ void __init reserve_crashkernel(void) memory_limit); } -lmb_reserve (crashk_res.start, size); +printk (KERN_INFO "Reserving %ldMB of memory at %ldMB " +"for crashkernel (System RAM: %ldMB)\n" , + (unsigned long )(crash_size >>20 ), + (unsigned long )(crashk_res.start >>20 ), + (unsigned long )(lmb_phys_mem_size () >>20 )); + +lmb_reserve (crashk_res.start, crash_size); }int overlaps_crashkernel (unsigned long start,unsigned long size)
Andrew Morton
maintainerRe: [patch 5/7] Use extended crashkernel command line on ppc64
Quoted 4 lines · On Tue, 25 Sep 2007 20:23:02 +0200 Bernhard Walle <[email protected]> wrote:
This patch adapts the ppc64 code to use the generic
parse_crashkernel()function introduced in the generic patch of that series.
I really don't like to see patches get a wholesale replacement, especially when they've been looked at by a few people and have had some testing, etc. It's not possible to see what changed and people need to re-review stuff they've already reviewed, etc.
So I almost always undo this mess, turn the patches back into incremental ones, see what pops out.
This patch is actually:
arch/powerpc/kernel/machine_kexec.c renamed from arch/powerpc/kernel/machine_kexec.c~use-extended-crashkernel-command-line-on-ppc64-update +1 -1
@@ -63,7 +63,7 @@ NORET_TYPE void machine_kexec(struct kimvoid __init reserve_crashkernel (void ) { -unsigned long long crash_size =0 , crash_base; +unsigned long long crash_size, crash_base;int ret;/* this is necessary because of lmb_phys_mem_size() */
_
which I suspect will now create a compiler warning.
unsigned long long crash_size, crash_base;
int ret;
/* this is necessary because of lmb_phys_mem_size() */
lmb_analyze();
/* use common parsing */
ret = parse_crashkernel(boot_command_line, lmb_phys_mem_size(),
&crash_size, &crash_base);
if (ret == 0 && crash_size > 0) {
if (crash_base == 0)
crash_base = KDUMP_KERNELBASE;
crashk_res.start = crash_base;
} else {
/* handle the device tree */
crash_size = crashk_res.end - crashk_res.start + 1;
}
if (crash_size == 0)
return;If so, the use of uninitialized_var() would be better than the unneeded initialization-to-zero.
_______________________________________________
kexec mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/kexec
Re: [patch 5/7] Use extended crashkernel command line on ppc64
Hi,
Quoted 9 lines · Andrew Morton
Quoted 2 lines · On Tue, 25 Sep 2007 20:23:02 +0200 Bernhard Walle <[email protected]> wrote:
This patch adapts the ppc64 code to use the generic
parse_crashkernel()function introduced in the generic patch of that series.I really don't like to see patches get a wholesale replacement, especially when they've been looked at by a few people and have had some testing, etc. It's not possible to see what changed and people need to re-review stuff they've already reviewed, etc.
Sorry, I didn't know that this is the preferred way. I just thought it's more clear if the patches are replaced if the patch set is finally pushed into Linus' tree.
Quoted 15 lines
diff -puN
arch/powerpc/kernel/machine_kexec.c~use-extended-crashkernel-command-line-on-ppc64-updatearch/powerpc/kernel/machine_kexec.carch/powerpc/kernel/machine_kexec.c renamed from arch/powerpc/kernel/machine_kexec.c~use-extended-crashkernel-command-line-on-ppc64-update +1
-1@@ -63,7 +63,7 @@ NORET_TYPE void machine_kexec(struct kimvoid __init reserve_crashkernel (void ) { -unsigned long long crash_size =0 , crash_base; +unsigned long long crash_size, crash_base;int ret;/* this is necessary because of lmb_phys_mem_size() */ _
Quoted 1 line
which I suspect will now create a compiler warning.
No, it doesn't. I just verified this.
CHK include/linux/compile.h CC arch/powerpc/kernel/machine_kexec.o LD arch/powerpc/kernel/built-in.o
Thanks,
Bernhard
Signature · 4 lines
Bernhard Walle <[email protected]>, Architecture Maintenance SUSE LINUX Products GmbH, Nürnberg, Germany GF: Markus Rex, HRB 16746 (AG Nürnberg) OpenPGP: F61F 34CC 09CA FB82 C9F6 BA4B 8865 3696 DDAF 6454
PATCH 6/7 Use extended crashkernel command line on sh
In reply to Bernhard WalleThis patch removes the crashkernel parsing from arch/sh/kernel/machine_kexec.c and calls the generic function, introduced in the generic patch, in setup_bootmem_allocator().
This is necessary because the amount of System RAM must be known in this function now because of the new syntax.
NOTE: Due to the lack of a SH processor, this patch is untested (and uncompiled). Because the code in that area is quite similar as i386/x86_64 (contrary to PPC and IA64), it should compile and work. However, if someone of the SH people could test for me and provide feedback, that would be very nice.
--- arch/sh/kernel/machine_kexec.c | 21 --------------------- arch/sh/kernel/setup.c | 38 +++++++++++++++++++++++++++++++++----- 2 files changed, 33 insertions(+), 26 deletions(-)
arch/sh/kernel/machine_kexec.c +0 -21
@@ -104,24 +104,3 @@ NORET_TYPE void machine_kexec(struct kim (*rnk)(page_list, reboot_code_buffer, image->start, vbr_reg); } -/* crashkernel=size@addr specifies the location to reserve for -* a crash kernel. By reserving this memory we guarantee -* that linux never sets it up as a DMA target. -* Useful for holding code to do something appropriate -* after a kernel panic. -*/ -static int __init parse_crashkernel (char *arg) -{ -unsigned long size, base; - size =memparse (arg, &arg); -if (*arg =='@' ) { - base =memparse (arg+1 , &arg); -/* FIXME: Do I want a sanity check -* to validate the memory range? -*/ - crashk_res.start = base; - crashk_res.end = base + size -1 ; - } -return 0 ; -} -early_param ("crashkernel" , parse_crashkernel);
arch/sh/kernel/setup.c +33 -5
@@ -128,6 +128,37 @@ static void __init register_bootmem_low_free_bootmem (PFN_PHYS (curr_pfn),PFN_PHYS (pages)); } +#ifdef CONFIG_KEXEC +static void __init reserve_crashkernel (void ) +{ +unsigned long long free_mem; +unsigned long long crash_size, crash_base; +int ret; + + free_mem = ((unsigned long long )max_low_pfn - min_low_pfn) <<PAGE_SHIFT ; + + ret =parse_crashkernel (boot_command_line, free_mem, + &crash_size, &crash_base); +if (ret ==0 && crash_size) { +if (crash_base >0 ) { +printk (KERN_INFO "Reserving %ldMB of memory at %ldMB " +"for crashkernel (System RAM: %ldMB)\n" , + (unsigned long )(crash_size >>20 ), + (unsigned long )(crash_base >>20 ), + (unsigned long )(free_mem >>20 )); + crashk_res.start = crash_base; + crashk_res.end = crash_base + crash_size -1 ; +reserve_bootmem (crash_base, crash_size); + }else +printk (KERN_INFO "crashkernel reservation failed - " +"you have to specify a base address\n" ); + } +} +#else +static inline void __init reserve_crashkernel (void ) +{} +#endif +void __init setup_bootmem_allocator (unsigned long free_pfn) {unsigned long bootmap_size;
@@ -189,11 +220,8 @@ void __init setup_bootmem_allocator(unsi } }#endif -#ifdef CONFIG_KEXEC -if (crashk_res.start != crashk_res.end) -reserve_bootmem (crashk_res.start, - crashk_res.end - crashk_res.start +1 ); -#endif + +reserve_crashkernel (); }#ifndef CONFIG_NEED_MULTIPLE_NODES
PATCH 7/7 Add documentation for extended crashkernel syntax
In reply to Bernhard WalleThis adds the documentation for the extended crashkernel syntax into Documentation/kdump/kdump.txt.
--- Documentation/kdump/kdump.txt | 26 ++++++++++++++++++++++++++ Documentation/kernel-parameters.txt | 7 +++++++ 2 files changed, 33 insertions(+)
Documentation/kdump/kdump.txt +26 -0
@@ -231,6 +231,32 @@ Dump-capture kernel config options (Arch any space below the alignment point will be wasted. +Extended crashkernel syntax +=========================== + +While the "crashkernel=size[@offset]" syntax is sufficient for most +configurations, sometimes it's handy to have the reserved memory dependent +on the value of System RAM -- that's mostly for distributors that pre-setup +the kernel command line to avoid a unbootable system after some memory has +been removed from the machine. + +The syntax is: + + crashkernel=<range1>:<size1>[,<range2>:<size2>,...][@offset] + range=start-[end] + +For example: + + crashkernel=512M-2G:64M,2G-:128M + +This would mean: + + 1) if the RAM is smaller than 512M, then don't reserve anything + (this is the "rescue" case) + 2) if the RAM size is between 512M and 2G, then reserve 64M + 3) if the RAM size is larger than 2G, then reserve 128M + + Boot into System Kernel =======================
Documentation/kernel-parameters.txt +7 -0
@@ -495,6 +495,13 @@ and is between 256 and 4096 characters. [KNL] Reserve a chunk of physical memory to hold a kernel to switch to with kexec on panic. + crashkernel=range1:size1[,range2:size2,...][@offset] + [KNL] Same as above, but depends on the memory + in the running system. The syntax of range is + start-[end] where start and end are both + a memory unit (amount[KMG]). See also + Documentation/kdump/kdump.txt for a example. + cs4232= [HW,OSS] Format: <io>,<irq>,<dma>,<dma2>,<mpuio>,<mpuirq>