[PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[WANG Chao]

We are using memset() to improve performance when creating 1st and 2nd
bitmap. After doing round up the pfn_start and round down pfn_end, it's
possible that pfn_start_roundup is greater than pfn_end_round. A segment
fault could happen in that case because memset is taking roughly the
value of (pfn_end_round << 3 - pfn_start_roundup << 3 ), which is
negative, as its third argument.

So we can skip the memset if start is greater than end. It's safe
because we will set bit for the round up part and also round down part.

Actually this happens on my EFI virtual machine:

cat /proc/iomem:
00000000-00000fff : reserved
00001000-0009ffff : System RAM
000a0000-000bffff : PCI Bus 0000:00
000f0000-000fffff : System ROM
00100000-3d162017 : System RAM
  01000000-015cab9b : Kernel code
  015cab9c-019beb3f : Kernel data
  01b4f000-01da9fff : Kernel bss
  30000000-37ffffff : Crash kernel
3d162018-3d171e57 : System RAM
3d171e58-3d172017 : System RAM
3d172018-3d17ae57 : System RAM
3d17ae58-3dc10fff : System RAM
3dc11000-3dc18fff : reserved
3dc19000-3dc41fff : System RAM
3dc42000-3ddcefff : reserved
3ddcf000-3f7fefff : System RAM
3f7ff000-3f856fff : reserved
[..]

gdb ./makedumpfile core
(gdb) bt full
[..]
 #1  0x000000000042775d in create_1st_bitmap_cyclic () at makedumpfile.c:4543
        i = 0x5
        pfn = 0x3d190
        phys_start = 0x3d18ee58
        phys_end = 0x3d18f018
        pfn_start = 0x3d18e
        pfn_end = 0x3d18f
        pfn_start_roundup = 0x3d190
        pfn_end_round = 0x3d188
        pfn_start_byte = 0x7a32
        pfn_end_byte = 0x7a31
[..]
(gdb) list makedumpfile.c:4543
4538					return FALSE;
4539
4540			pfn_start_byte = (pfn_start_roundup - info->cyclic_start_pfn) >> 3;
4541			pfn_end_byte = (pfn_end_round - info->cyclic_start_pfn) >> 3;
4542
4543			memset(info->partial_bitmap2 + pfn_start_byte,
4544			       0xff,
4545			       pfn_end_byte - pfn_start_byte);
4546
4547			for (pfn = pfn_end_round; pfn < pfn_end; ++pfn)

Signed-off-by: WANG Chao <chaowang@redhat.com>
---
 makedumpfile.c | 18 +++++++++++-------
 1 file changed, 11 insertions(+), 7 deletions(-)

diff --git a/makedumpfile.c b/makedumpfile.c
index 23251a1..ef08d91 100644
--- a/makedumpfile.c
+++ b/makedumpfile.c
@@ -4435,11 +4435,13 @@ create_1st_bitmap_cyclic()
 		pfn_start_byte = (pfn_start_roundup - info->cyclic_start_pfn) >> 3;
 		pfn_end_byte = (pfn_end_round - info->cyclic_start_pfn) >> 3;
 
-		memset(info->partial_bitmap1 + pfn_start_byte,
-		       0xff,
-		       pfn_end_byte - pfn_start_byte);
+		if (pfn_start_byte < pfn_end_byte) {
+			memset(info->partial_bitmap1 + pfn_start_byte,
+			       0xff,
+			       pfn_end_byte - pfn_start_byte);
 
-		pfn_bitmap1 += (pfn_end_byte - pfn_start_byte) * BITPERBYTE;
+			pfn_bitmap1 += (pfn_end_byte - pfn_start_byte) * BITPERBYTE;
+		}
 
 		for (pfn = pfn_end_round; pfn < pfn_end; pfn++) {
 			if (set_bit_on_1st_bitmap(pfn))
@@ -4540,9 +4542,11 @@ initialize_2nd_bitmap_cyclic(void)
 		pfn_start_byte = (pfn_start_roundup - info->cyclic_start_pfn) >> 3;
 		pfn_end_byte = (pfn_end_round - info->cyclic_start_pfn) >> 3;
 
-		memset(info->partial_bitmap2 + pfn_start_byte,
-		       0xff,
-		       pfn_end_byte - pfn_start_byte);
+		if (pfn_start_byte < pfn_end_byte) {
+			memset(info->partial_bitmap2 + pfn_start_byte,
+			       0xff,
+			       pfn_end_byte - pfn_start_byte);
+		}
 
 		for (pfn = pfn_end_round; pfn < pfn_end; ++pfn)
 			if (!set_bit_on_2nd_bitmap_for_kernel(pfn))
-- 
1.8.4.2


_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

Re: [PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[HATAYAMA Daisuke]

(2013/12/18 22:34), WANG Chao wrote:
> We are using memset() to improve performance when creating 1st and 2nd
> bitmap. After doing round up the pfn_start and round down pfn_end, it's
> possible that pfn_start_roundup is greater than pfn_end_round. A segment
> fault could happen in that case because memset is taking roughly the
> value of (pfn_end_round << 3 - pfn_start_roundup << 3 ), which is
> negative, as its third argument.
>
> So we can skip the memset if start is greater than end. It's safe
> because we will set bit for the round up part and also round down part.
>
> Actually this happens on my EFI virtual machine:
>
> cat /proc/iomem:
> 00000000-00000fff : reserved
> 00001000-0009ffff : System RAM
> 000a0000-000bffff : PCI Bus 0000:00
> 000f0000-000fffff : System ROM
> 00100000-3d162017 : System RAM
>    01000000-015cab9b : Kernel code
>    015cab9c-019beb3f : Kernel data
>    01b4f000-01da9fff : Kernel bss
>    30000000-37ffffff : Crash kernel
> 3d162018-3d171e57 : System RAM
> 3d171e58-3d172017 : System RAM
> 3d172018-3d17ae57 : System RAM
> 3d17ae58-3dc10fff : System RAM
> 3dc11000-3dc18fff : reserved
> 3dc19000-3dc41fff : System RAM
> 3dc42000-3ddcefff : reserved
> 3ddcf000-3f7fefff : System RAM
> 3f7ff000-3f856fff : reserved
> [..]
>
> gdb ./makedumpfile core
> (gdb) bt full
> [..]
>   #1  0x000000000042775d in create_1st_bitmap_cyclic () at makedumpfile.c:4543
>          i = 0x5
>          pfn = 0x3d190
>          phys_start = 0x3d18ee58
>          phys_end = 0x3d18f018
>          pfn_start = 0x3d18e
>          pfn_end = 0x3d18f
>          pfn_start_roundup = 0x3d190
>          pfn_end_round = 0x3d188
>          pfn_start_byte = 0x7a32
>          pfn_end_byte = 0x7a31
> [..]
> (gdb) list makedumpfile.c:4543
> 4538					return FALSE;
> 4539
> 4540			pfn_start_byte = (pfn_start_roundup - info->cyclic_start_pfn) >> 3;
> 4541			pfn_end_byte = (pfn_end_round - info->cyclic_start_pfn) >> 3;
> 4542
> 4543			memset(info->partial_bitmap2 + pfn_start_byte,
> 4544			       0xff,
> 4545			       pfn_end_byte - pfn_start_byte);
> 4546
> 4547			for (pfn = pfn_end_round; pfn < pfn_end; ++pfn)
>
> Signed-off-by: WANG Chao <chaowang@redhat.com>
> ---
>   makedumpfile.c | 18 +++++++++++-------
>   1 file changed, 11 insertions(+), 7 deletions(-)
>
> diff --git a/makedumpfile.c b/makedumpfile.c
> index 23251a1..ef08d91 100644
> --- a/makedumpfile.c
> +++ b/makedumpfile.c
> @@ -4435,11 +4435,13 @@ create_1st_bitmap_cyclic()
>   		pfn_start_byte = (pfn_start_roundup - info->cyclic_start_pfn) >> 3;
>   		pfn_end_byte = (pfn_end_round - info->cyclic_start_pfn) >> 3;
>
> -		memset(info->partial_bitmap1 + pfn_start_byte,
> -		       0xff,
> -		       pfn_end_byte - pfn_start_byte);
> +		if (pfn_start_byte < pfn_end_byte) {
> +			memset(info->partial_bitmap1 + pfn_start_byte,
> +			       0xff,
> +			       pfn_end_byte - pfn_start_byte);
>
> -		pfn_bitmap1 += (pfn_end_byte - pfn_start_byte) * BITPERBYTE;
> +			pfn_bitmap1 += (pfn_end_byte - pfn_start_byte) * BITPERBYTE;
> +		}
>
>   		for (pfn = pfn_end_round; pfn < pfn_end; pfn++) {
>   			if (set_bit_on_1st_bitmap(pfn))
> @@ -4540,9 +4542,11 @@ initialize_2nd_bitmap_cyclic(void)
>   		pfn_start_byte = (pfn_start_roundup - info->cyclic_start_pfn) >> 3;
>   		pfn_end_byte = (pfn_end_round - info->cyclic_start_pfn) >> 3;
>
> -		memset(info->partial_bitmap2 + pfn_start_byte,
> -		       0xff,
> -		       pfn_end_byte - pfn_start_byte);
> +		if (pfn_start_byte < pfn_end_byte) {
> +			memset(info->partial_bitmap2 + pfn_start_byte,
> +			       0xff,
> +			       pfn_end_byte - pfn_start_byte);
> +		}
>
>   		for (pfn = pfn_end_round; pfn < pfn_end; ++pfn)
>   			if (!set_bit_on_2nd_bitmap_for_kernel(pfn))
>

Acked-by: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com>

Also, I'm interested in the memory map passed to from EFI in that

> cat /proc/iomem:
> 00000000-00000fff : reserved
> 00001000-0009ffff : System RAM
> 000a0000-000bffff : PCI Bus 0000:00
> 000f0000-000fffff : System ROM
> 00100000-3d162017 : System RAM
>    01000000-015cab9b : Kernel code
>    015cab9c-019beb3f : Kernel data
>    01b4f000-01da9fff : Kernel bss
>    30000000-37ffffff : Crash kernel
> 3d162018-3d171e57 : System RAM
> 3d171e58-3d172017 : System RAM
> 3d172018-3d17ae57 : System RAM
> 3d17ae58-3dc10fff : System RAM

this part is consecutive but somehow is divided into 4 entries.
You called your environment as ``EFI virtual machine'', could you tell
me precisely what it mean? qemu/KVM or VMware guest system? I do want
to understand how this kind of memory map was created. I think this
kind of memory mapping is odd and I guess this is caused by the fact
that the system is a virtual environment.

And for Vivek, this case is a concrete example of multiple RAM entries
appearing in a single page I suspected in the mmap failure patch,
although these entries are consecutive in physical address and can be
represented by a single entry by merging them in a single entry. But
then it seems to me that there could be more odd case that multiple
RAM entries but not consecutive. I again think this should be addressed
in the patch for the mmap failure issue. How do you think?

> 3dc11000-3dc18fff : reserved
> 3dc19000-3dc41fff : System RAM
> 3dc42000-3ddcefff : reserved
> 3ddcf000-3f7fefff : System RAM
> 3f7ff000-3f856fff : reserved
> [..]

-- 
Thanks.
HATAYAMA, Daisuke


_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

Re: [PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[Dave Young]

> Also, I'm interested in the memory map passed to from EFI in that
> 
> >cat /proc/iomem:
> >00000000-00000fff : reserved
> >00001000-0009ffff : System RAM
> >000a0000-000bffff : PCI Bus 0000:00
> >000f0000-000fffff : System ROM
> >00100000-3d162017 : System RAM
> >   01000000-015cab9b : Kernel code
> >   015cab9c-019beb3f : Kernel data
> >   01b4f000-01da9fff : Kernel bss
> >   30000000-37ffffff : Crash kernel
> >3d162018-3d171e57 : System RAM
> >3d171e58-3d172017 : System RAM
> >3d172018-3d17ae57 : System RAM
> >3d17ae58-3dc10fff : System RAM
> 
> this part is consecutive but somehow is divided into 4 entries.
> You called your environment as ``EFI virtual machine'', could you tell
> me precisely what it mean? qemu/KVM or VMware guest system? I do want
> to understand how this kind of memory map was created. I think this
> kind of memory mapping is odd and I guess this is caused by the fact
> that the system is a virtual environment.

This is not specific to EFI machine, it's the reserved setup_data regions
They happened to be continous but they do not have to be continuous.

> 
> And for Vivek, this case is a concrete example of multiple RAM entries
> appearing in a single page I suspected in the mmap failure patch,
> although these entries are consecutive in physical address and can be
> represented by a single entry by merging them in a single entry. But
> then it seems to me that there could be more odd case that multiple
> RAM entries but not consecutive. I again think this should be addressed
> in the patch for the mmap failure issue. How do you think?

They are different problems, the previous mmap bug is for cross page regions
with different page flags.

> 
> >3dc11000-3dc18fff : reserved
> >3dc19000-3dc41fff : System RAM
> >3dc42000-3ddcefff : reserved
> >3ddcf000-3f7fefff : System RAM
> >3f7ff000-3f856fff : reserved
> >[..]
> 

_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

Re: [PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[Atsushi Kumagai]

On 2013/12/20 10:09:23, kexec <kexec-bounces@lists.infradead.org> wrote:
> (2013/12/18 22:34), WANG Chao wrote:
> > We are using memset() to improve performance when creating 1st and 2nd
> > bitmap. After doing round up the pfn_start and round down pfn_end, it's
> > possible that pfn_start_roundup is greater than pfn_end_round. A segment
> > fault could happen in that case because memset is taking roughly the
> > value of (pfn_end_round << 3 - pfn_start_roundup << 3 ), which is
> > negative, as its third argument.
> >
> > So we can skip the memset if start is greater than end. It's safe
> > because we will set bit for the round up part and also round down part.
> >
> > Actually this happens on my EFI virtual machine:
> >
> > cat /proc/iomem:
> > 00000000-00000fff : reserved
> > 00001000-0009ffff : System RAM
> > 000a0000-000bffff : PCI Bus 0000:00
> > 000f0000-000fffff : System ROM
> > 00100000-3d162017 : System RAM
> >    01000000-015cab9b : Kernel code
> >    015cab9c-019beb3f : Kernel data
> >    01b4f000-01da9fff : Kernel bss
> >    30000000-37ffffff : Crash kernel
> > 3d162018-3d171e57 : System RAM
> > 3d171e58-3d172017 : System RAM
> > 3d172018-3d17ae57 : System RAM
> > 3d17ae58-3dc10fff : System RAM
> > 3dc11000-3dc18fff : reserved
> > 3dc19000-3dc41fff : System RAM
> > 3dc42000-3ddcefff : reserved
> > 3ddcf000-3f7fefff : System RAM
> > 3f7ff000-3f856fff : reserved
> > [..]
> >
> > gdb ./makedumpfile core
> > (gdb) bt full
> > [..]
> >   #1  0x000000000042775d in create_1st_bitmap_cyclic () at makedumpfile.c:4543
> >          i = 0x5
> >          pfn = 0x3d190
> >          phys_start = 0x3d18ee58
> >          phys_end = 0x3d18f018
> >          pfn_start = 0x3d18e
> >          pfn_end = 0x3d18f
> >          pfn_start_roundup = 0x3d190
> >          pfn_end_round = 0x3d188
> >          pfn_start_byte = 0x7a32
> >          pfn_end_byte = 0x7a31
> > [..]
> > (gdb) list makedumpfile.c:4543
> > 4538					return FALSE;
> > 4539
> > 4540			pfn_start_byte = (pfn_start_roundup - info->cyclic_start_pfn) >> 3;
> > 4541			pfn_end_byte = (pfn_end_round - info->cyclic_start_pfn) >> 3;
> > 4542
> > 4543			memset(info->partial_bitmap2 + pfn_start_byte,
> > 4544			       0xff,
> > 4545			       pfn_end_byte - pfn_start_byte);
> > 4546
> > 4547			for (pfn = pfn_end_round; pfn < pfn_end; ++pfn)
> >
> > Signed-off-by: WANG Chao <chaowang@redhat.com>
> > ---
> >   makedumpfile.c | 18 +++++++++++-------
> >   1 file changed, 11 insertions(+), 7 deletions(-)
> >
> > diff --git a/makedumpfile.c b/makedumpfile.c
> > index 23251a1..ef08d91 100644
> > --- a/makedumpfile.c
> > +++ b/makedumpfile.c
> > @@ -4435,11 +4435,13 @@ create_1st_bitmap_cyclic()
> >   		pfn_start_byte = (pfn_start_roundup - info->cyclic_start_pfn) >> 3;
> >   		pfn_end_byte = (pfn_end_round - info->cyclic_start_pfn) >> 3;
> >
> > -		memset(info->partial_bitmap1 + pfn_start_byte,
> > -		       0xff,
> > -		       pfn_end_byte - pfn_start_byte);
> > +		if (pfn_start_byte < pfn_end_byte) {
> > +			memset(info->partial_bitmap1 + pfn_start_byte,
> > +			       0xff,
> > +			       pfn_end_byte - pfn_start_byte);
> >
> > -		pfn_bitmap1 += (pfn_end_byte - pfn_start_byte) * BITPERBYTE;
> > +			pfn_bitmap1 += (pfn_end_byte - pfn_start_byte) * BITPERBYTE;
> > +		}
> >
> >   		for (pfn = pfn_end_round; pfn < pfn_end; pfn++) {
> >   			if (set_bit_on_1st_bitmap(pfn))
> > @@ -4540,9 +4542,11 @@ initialize_2nd_bitmap_cyclic(void)
> >   		pfn_start_byte = (pfn_start_roundup - info->cyclic_start_pfn) >> 3;
> >   		pfn_end_byte = (pfn_end_round - info->cyclic_start_pfn) >> 3;
> >
> > -		memset(info->partial_bitmap2 + pfn_start_byte,
> > -		       0xff,
> > -		       pfn_end_byte - pfn_start_byte);
> > +		if (pfn_start_byte < pfn_end_byte) {
> > +			memset(info->partial_bitmap2 + pfn_start_byte,
> > +			       0xff,
> > +			       pfn_end_byte - pfn_start_byte);
> > +		}
> >
> >   		for (pfn = pfn_end_round; pfn < pfn_end; ++pfn)
> >   			if (!set_bit_on_2nd_bitmap_for_kernel(pfn))
> >
> 
> Acked-by: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com>

Thanks, I'll merge this patch into v1.5.6.

Atsushi Kumagai

_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

Re: [PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[HATAYAMA Daisuke]

(2013/12/20 11:17), Dave Young wrote:
>> Also, I'm interested in the memory map passed to from EFI in that
>>
>>> cat /proc/iomem:
>>> 00000000-00000fff : reserved
>>> 00001000-0009ffff : System RAM
>>> 000a0000-000bffff : PCI Bus 0000:00
>>> 000f0000-000fffff : System ROM
>>> 00100000-3d162017 : System RAM
>>>    01000000-015cab9b : Kernel code
>>>    015cab9c-019beb3f : Kernel data
>>>    01b4f000-01da9fff : Kernel bss
>>>    30000000-37ffffff : Crash kernel
>>> 3d162018-3d171e57 : System RAM
>>> 3d171e58-3d172017 : System RAM
>>> 3d172018-3d17ae57 : System RAM
>>> 3d17ae58-3dc10fff : System RAM
>>
>> this part is consecutive but somehow is divided into 4 entries.
>> You called your environment as ``EFI virtual machine'', could you tell
>> me precisely what it mean? qemu/KVM or VMware guest system? I do want
>> to understand how this kind of memory map was created. I think this
>> kind of memory mapping is odd and I guess this is caused by the fact
>> that the system is a virtual environment.
>
> This is not specific to EFI machine, it's the reserved setup_data regions
> They happened to be continous but they do not have to be continuous.
>

Thanks for pointing out that. I've just read Documentation/x86/boot.txt
and parse_setup_data().

But I don't understand well why these regions are divided as these.
I guess kernel divides the System RAM this way and the memory map first
passed to by EFI is all page aligned, right?

Also, looking at parse_setup_data(), currently handled data in setup_data
interface is extended e820 entries and dtb case only.

                 switch (data_type) {
                 case SETUP_E820_EXT:
                         parse_e820_ext(pa_data, data_len);
                         break;
                 case SETUP_DTB:
                         add_dtb(pa_data);
                         break;
                 default:
                         break;
                 }

Is it right that this kind of memory map doesn't occur as long as either
of information is passed to via setup_data? IOW, is this necessary
information?

>>
>> And for Vivek, this case is a concrete example of multiple RAM entries
>> appearing in a single page I suspected in the mmap failure patch,
>> although these entries are consecutive in physical address and can be
>> represented by a single entry by merging them in a single entry. But
>> then it seems to me that there could be more odd case that multiple
>> RAM entries but not consecutive. I again think this should be addressed
>> in the patch for the mmap failure issue. How do you think?
>
> They are different problems, the previous mmap bug is for cross page regions
> with different page flags.
>

I understand that. What I think problem here is the case where multiple
System RAM entries appear in a single page. In the above memory map, they
are 3d171000, 3d172000 and 3d17a000. My fixing patch is to copy fractional
pages in the 2nd kernel in order to make it possible to mmap without affecting
non-System RAM area as much as possible, and then if there is this kind of
System RAM entries, we need to use the same page in the 2nd kernel for
different System RAM entries that shares the same page in the 1st kernel. This
needs a little additional processing and we want to keep implementation as
simple as possible as long as there's no such system in real world. However,
I'm surprised to see the memory mapping above.

-- 
Thanks.
HATAYAMA, Daisuke


_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

Re: [PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[Dave Young]

On 12/20/13 at 05:49pm, HATAYAMA Daisuke wrote:
> (2013/12/20 11:17), Dave Young wrote:
> >>Also, I'm interested in the memory map passed to from EFI in that
> >>
> >>>cat /proc/iomem:
> >>>00000000-00000fff : reserved
> >>>00001000-0009ffff : System RAM
> >>>000a0000-000bffff : PCI Bus 0000:00
> >>>000f0000-000fffff : System ROM
> >>>00100000-3d162017 : System RAM
> >>>   01000000-015cab9b : Kernel code
> >>>   015cab9c-019beb3f : Kernel data
> >>>   01b4f000-01da9fff : Kernel bss
> >>>   30000000-37ffffff : Crash kernel
> >>>3d162018-3d171e57 : System RAM
> >>>3d171e58-3d172017 : System RAM
> >>>3d172018-3d17ae57 : System RAM
> >>>3d17ae58-3dc10fff : System RAM
> >>
> >>this part is consecutive but somehow is divided into 4 entries.
> >>You called your environment as ``EFI virtual machine'', could you tell
> >>me precisely what it mean? qemu/KVM or VMware guest system? I do want
> >>to understand how this kind of memory map was created. I think this
> >>kind of memory mapping is odd and I guess this is caused by the fact
> >>that the system is a virtual environment.
> >
> >This is not specific to EFI machine, it's the reserved setup_data regions
> >They happened to be continous but they do not have to be continuous.
> >
> 
> Thanks for pointing out that. I've just read Documentation/x86/boot.txt
> and parse_setup_data().
> 
> But I don't understand well why these regions are divided as these.
> I guess kernel divides the System RAM this way and the memory map first
> passed to by EFI is all page aligned, right?

setup_data are passed as a link list by boot loader, each node is a block
of memory, there could be many different setup_data type.

> 
> Also, looking at parse_setup_data(), currently handled data in setup_data
> interface is extended e820 entries and dtb case only.
> 
>                 switch (data_type) {
>                 case SETUP_E820_EXT:
>                         parse_e820_ext(pa_data, data_len);
>                         break;
>                 case SETUP_DTB:
>                         add_dtb(pa_data);
>                         break;
>                 default:
>                         break;
>                 }
> 
> Is it right that this kind of memory map doesn't occur as long as either
> of information is passed to via setup_data? IOW, is this necessary
> information?

If bootloader does not pass it, there will be no such mem ranges in /proc/iomem.

> 
> >>
> >>And for Vivek, this case is a concrete example of multiple RAM entries
> >>appearing in a single page I suspected in the mmap failure patch,
> >>although these entries are consecutive in physical address and can be
> >>represented by a single entry by merging them in a single entry. But
> >>then it seems to me that there could be more odd case that multiple
> >>RAM entries but not consecutive. I again think this should be addressed
> >>in the patch for the mmap failure issue. How do you think?
> >
> >They are different problems, the previous mmap bug is for cross page regions
> >with different page flags.
> >
> 
> I understand that. What I think problem here is the case where multiple
> System RAM entries appear in a single page. In the above memory map, they
> are 3d171000, 3d172000 and 3d17a000. My fixing patch is to copy fractional
> pages in the 2nd kernel in order to make it possible to mmap without affecting
> non-System RAM area as much as possible, and then if there is this kind of
> System RAM entries, we need to use the same page in the 2nd kernel for
> different System RAM entries that shares the same page in the 1st kernel. This
> needs a little additional processing and we want to keep implementation as
> simple as possible as long as there's no such system in real world. However,
> I'm surprised to see the memory mapping above.

These ranges are "system ram" of type E820_RESERVED_KERN, please see below:
arch/x86/kernel/setup.c: e820_reserve_setup_data()

Thanks
Dave

_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

Re: [PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[Lisa Mitchell]

On Fri, 2013-12-20 at 14:13 +0000, Vivek Goyal wrote:
> On Fri, Dec 20, 2013 at 10:08:08AM +0900, HATAYAMA Daisuke wrote:
> 
> [..]
> > 
> > >cat /proc/iomem:
> > >00000000-00000fff : reserved
> > >00001000-0009ffff : System RAM
> > >000a0000-000bffff : PCI Bus 0000:00
> > >000f0000-000fffff : System ROM
> > >00100000-3d162017 : System RAM
> > >   01000000-015cab9b : Kernel code
> > >   015cab9c-019beb3f : Kernel data
> > >   01b4f000-01da9fff : Kernel bss
> > >   30000000-37ffffff : Crash kernel
> > >3d162018-3d171e57 : System RAM
> > >3d171e58-3d172017 : System RAM
> > >3d172018-3d17ae57 : System RAM
> > >3d17ae58-3dc10fff : System RAM
> > 
> > this part is consecutive but somehow is divided into 4 entries.
> > You called your environment as ``EFI virtual machine'', could you tell
> > me precisely what it mean? qemu/KVM or VMware guest system? I do want
> > to understand how this kind of memory map was created. I think this
> > kind of memory mapping is odd and I guess this is caused by the fact
> > that the system is a virtual environment.
> > 
> > And for Vivek, this case is a concrete example of multiple RAM entries
> > appearing in a single page I suspected in the mmap failure patch,
> > although these entries are consecutive in physical address and can be
> > represented by a single entry by merging them in a single entry. But
> > then it seems to me that there could be more odd case that multiple
> > RAM entries but not consecutive. I again think this should be addressed
> > in the patch for the mmap failure issue. How do you think?
> 
> Hi Hatayama,
> 
> This indeed looks very odd. See if a very small number of systems have it,
> the only thing we will do is allocate extra page in second kernel for
> a memory range. It will not make mmap() fail. So it is just a matter of
> optimization.
> 
> Given the fact I have not seen many systems with this anomaly, I am not
> too worried about it even if you don't this optimization in your patch
> series. We can always take care of it later if need be.
> 
> At the same time, if you feel strongly about it and want to fix it in
> same patch series, I don't mind.
> 
> Thanks
> Vivek

Did I get this same segmentation fault?   It happened a few times on a
3.10 based kernel on a large EFI based system , but then hasn't repeated
in further testing on this machine. This machine had no virtualization
active. Here is partial console log of dump process:  

=============================================================================

kdump: dump target is /dev/mapper/mpathc3
kdump: saving to /sysroot//var/crash/127.0.0.1-2013.11.11-09:58:20/
kdump: saving vmcore-dmesg.txt
kdump: saving vmcore-dmesg.txt complete
kdump: saving vmcore
[   46.017827] scsi 20:0:0:0: Direct-Access              Geek Squad
8192 PQ: 0 ANSI: 0 CCS
[   46.027849] scsi 20:0:0:0: alua: supports implicit and explicit TPGS
[   46.035641] scsi 20:0:0:0: alua: No target port descriptors found
[   46.042491] scsi 20:0:0:0: alua: not attached
[   46.048152] sd 20:0:0:0: [sdak] 15663104 512-byte logical blocks:
(8.01 GB/7.46 GiB)
[   46.057892] sd 20:0:0:0: [sdak] Write Protect is off
[   46.064516] sd 20:0:0:0: [sdak] No Caching mode page present
[   46.070877] sd 20:0:0:0: [sdak] Assuming drive cache: write through
[   46.081639] sd 20:0:0:0: [sdak] No Caching mode page present
[   46.088035] sd 20:0:0:0: [sdak] Assuming drive cache: write through
[   46.107070]  sdak: sdak1 sdak2 sdak3
[   46.114636] sd 20:0:0:0: [sdak] No Caching mode page present
[   46.121021] sd 20:0:0:0: [sdak] Assuming drive cache: write through
[   46.128060] sd 20:0:0:0: [sdak] Attached SCSI removable disk

Excluding unnecessary pages        : [  0 %] 
Excluding unnecessary 
sary pages        : [  0 %] 
Excluding unnecessary pages        : [  0
%] 
Excluding unnecessary pages        : [  8 %] 
Excluding unnecessary
pages        : [ 15 %] 
Excluding unnecessary pages        : [ 21 %] [
53.800116] usb 4-1.6: device descriptor read/64, error -110

Excluding unnecessary pages        : [ 28 %] 
Excluding unnecessary
pages        : [ 34 %] 
Excluding unnecessary pages        : [ 41 %]
Excluding unnecessary pages        : [100 %] [   59.174433] scsi
18:0:0:0: Enclosure         HP       P2000 G3 FC      T240 PQ: 0 ANSI: 5
[   59.183866] scsi 18:0:0:0: alua: supports implicit TPGS
[   59.190190] scsi 18:0:0:0: alua: port group 01 rel port 05
[   59.197694] scsi 18:0:0:0: alua: transition timeout set to 60 seconds

......

[   76.968265] makedumpfile[1291]: segfault at 7f5ac0c39010 ip
00000000004297fd sp 00007fff691683a0 error 4 in makedumpfile[400000
+46000]
//lib/dracut/hooks/pre-pivot/9999-kdump.sh: line 88:  1291 Segmentation
fault   [   76.996926] sd 18:0:1:51: [sdao] Synchronizing SCSI cache
   $CORE_COLLECT[   77.003416] sd 18:0:1:50: [sdan] Synchronizing SCSI
cache
OR /proc/vmcore [   77.010991] sd 18:0:0:51: [sdam] Synchronizing SCSI
cache
$_mp/$KDUMP_PATH[   77.018590] sd 18:0:0:50: [sdal] Synchronizing SCSI
cache
/$HOST_IP-$DATED[   77.026611] sd 14:0:1:71: [sdaj] Synchronizing SCSI
cache
IR/vmcore-incomp[   77.033849] sd 14:0:1:70: [sdai] Synchronizing SCSI
cache
lete
kdump: sav[   77.041467] sd 14:0:0:71: [sdah] Synchronizing SCSI cache
ing vmcore faile[   77.049146] sd 14:0:0:70: [sdag] Synchronizing SCSI
cached
Rebooting.


=======================================================================

I do not have a current /proc/iomem output to go with the above.
However this dump was nr_cpus=8 during crashkernel boot, with a 3.10
kernel with Daisuke's version 4 of Daisuke Hatayama's patch to allow
multi-cpu crashkernel boot. 

I have the efi memory map displayed as Linux booted, before this dump if
that helps

0.000000] efi:  ACPI=0x73ffe000  ACPI 2.0=0x73ffe014  SMBIOS=0x72ef8000 
[    0.000000] efi: mem00: type=3, attr=0xf,
range=[0x0000000000000000-0x0000000000001000) (0MB)
[    0.000000] efi: mem01: type=2, attr=0xf,
range=[0x0000000000001000-0x0000000000004000) (0MB)
[    0.000000] efi: mem02: type=7, attr=0xf,
range=[0x0000000000004000-0x000000000008e000) (0MB)
[    0.000000] efi: mem03: type=0, attr=0xf,
range=[0x000000000008e000-0x0000000000090000) (0MB)
[    0.000000] efi: mem04: type=7, attr=0xf,
range=[0x0000000000090000-0x00000000000a0000) (0MB)
[    0.000000] efi: mem05: type=7, attr=0xf,
range=[0x0000000000100000-0x0000000001000000) (15MB)
[    0.000000] efi: mem06: type=2, attr=0xf,
range=[0x0000000001000000-0x0000000002268000) (18MB)
[    0.000000] efi: mem07: type=7, attr=0xf,
range=[0x0000000002268000-0x0000000010000000) (221MB)
[    0.000000] efi: mem08: type=3, attr=0xf,
range=[0x0000000010000000-0x0000000010066000) (0MB)
[    0.000000] efi: mem09: type=7, attr=0xf,
range=[0x0000000010066000-0x0000000029dfb000) (413MB)
[    0.000000] efi: mem10: type=2, attr=0xf,
range=[0x0000000029dfb000-0x0000000039ba0000) (253MB)
[    0.000000] efi: mem11: type=4, attr=0xf,
range=[0x0000000039ba0000-0x000000003e179000) (69MB)
[    0.000000] efi: mem12: type=3, attr=0xf,
range=[0x000000003e179000-0x000000003e55c000) (3MB)
[    0.000000] efi: mem13: type=4, attr=0xf,
range=[0x000000003e55c000-0x000000003e567000) (0MB)
[    0.000000] efi: mem14: type=3, attr=0xf,
range=[0x000000003e567000-0x000000003e6e9000) (1MB)
[    0.000000] efi: mem15: type=4, attr=0xf,
range=[0x000000003e6e9000-0x000000003e6f4000) (0MB)
[    0.000000] efi: mem16: type=3, attr=0xf,
range=[0x000000003e6f4000-0x000000003e88f000) (1MB)
[    0.000000] efi: mem17: type=4, attr=0xf,
range=[0x000000003e88f000-0x000000003e8f4000) (0MB)
[    0.000000] efi: mem18: type=3, attr=0xf,
range=[0x000000003e8f4000-0x000000003e9cf000) (0MB)
[    0.000000] efi: mem19: type=4, attr=0xf,
range=[0x000000003e9cf000-0x000000003e9d0000) (0MB)
[    0.000000] efi: mem20: type=3, attr=0xf,
range=[0x000000003e9d0000-0x000000003ee9d000) (4MB)
[    0.000000] efi: mem21: type=4, attr=0xf,
range=[0x000000003ee9d000-0x000000003ef9b000) (0MB)
[    0.000000] efi: mem22: type=0, attr=0xf,
range=[0x000000003ef9b000-0x000000003efab000) (0MB)
[    0.000000] efi: mem23: type=4, attr=0xf,
range=[0x000000003efab000-0x000000006ba1b000) (714MB)
[    0.000000] efi: mem24: type=10, attr=0xf,
range=[0x000000006ba1b000-0x000000006ca1b000) (16MB)
[    0.000000] efi: mem25: type=4, attr=0xf,
range=[0x000000006ca1b000-0x00000000709ff000) (63MB)
[    0.000000] efi: mem26: type=7, attr=0xf,
range=[0x00000000709ff000-0x0000000070a10000) (0MB)
[    0.000000] efi: mem27: type=2, attr=0xf,
range=[0x0000000070a10000-0x0000000070c34000) (2MB)
[    0.000000] efi: mem28: type=7, attr=0xf,
range=[0x0000000070c34000-0x0000000070c35000) (0MB)
[    0.000000] efi: mem29: type=2, attr=0xf,
range=[0x0000000070c35000-0x0000000070dff000) (1MB)
[    0.000000] efi: mem30: type=7, attr=0xf,
range=[0x0000000070dff000-0x0000000070fa9000) (1MB)
[    0.000000] efi: mem31: type=1, attr=0xf,
range=[0x0000000070fa9000-0x00000000711ff000) (2MB)
[    0.000000] efi: mem32: type=7, attr=0xf,
range=[0x00000000711ff000-0x000000007120e000) (0MB)
[    0.000000] efi: mem33: type=3, attr=0xf,
range=[0x000000007120e000-0x00000000721ff000) (15MB)
[    0.000000] efi: mem34: type=6, attr=0x800000000000000f,
range=[0x00000000721ff000-0x00000000725ff000) (4MB)
[    0.000000] efi: mem35: type=5, attr=0x800000000000000f,
range=[0x00000000725ff000-0x0000000072dff000) (8MB)
[    0.000000] efi: mem36: type=0, attr=0xf,
range=[0x0000000072dff000-0x0000000072eff000) (1MB)
[    0.000000] efi: mem37: type=10, attr=0xf,
range=[0x0000000072eff000-0x0000000073eff000) (16MB)
[    0.000000] efi: mem38: type=9, attr=0xf,
range=[0x0000000073eff000-0x0000000073fff000) (1MB)
[    0.000000] efi: mem39: type=4, attr=0xf,
range=[0x0000000073fff000-0x000000007c000000) (128MB)
[    0.000000] efi: mem40: type=7, attr=0xf,
range=[0x0000000100000000-0x0000001080000000) (63488MB)
[    0.000000] efi: mem41: type=7, attr=0xf,
range=[0x0000020000000000-0x0000021000000000) (65536MB)
[    0.000000] efi: mem42: type=7, attr=0xf,
range=[0x0000038000000000-0x0000039000000000) (65536MB)
[    0.000000] efi: mem43: type=7, attr=0xf,
range=[0x0000050000000000-0x0000051000000000) (65536MB)
[    0.000000] efi: mem44: type=7, attr=0xf,
range=[0x0000068000000000-0x0000069000000000) (65536MB)
[    0.000000] efi: mem45: type=7, attr=0xf,
range=[0x0000080000000000-0x0000081000000000) (65536MB)
[    0.000000] efi: mem46: type=7, attr=0xf,
range=[0x0000098000000000-0x0000099000000000) (65536MB)
[    0.000000] efi: mem47: type=7, attr=0xf,
range=[0x00000b0000000000-0x00000b1000000000) (65536MB)
[    0.000000] efi: mem48: type=7, attr=0xf,
range=[0x00000c8000000000-0x00000c9000000000) (65536MB)
[    0.000000] efi: mem49: type=7, attr=0xf,
range=[0x00000e0000000000-0x00000e1000000000) (65536MB)
[    0.000000] efi: mem50: type=7, attr=0xf,
range=[0x00000f8000000000-0x00000f9000000000) (65536MB)
[    0.000000] efi: mem51: type=7, attr=0xf,
range=[0x0000110000000000-0x0000111000000000) (65536MB)
[    0.000000] efi: mem52: type=7, attr=0xf,
range=[0x0000128000000000-0x0000129000000000) (65536MB)
[    0.000000] efi: mem53: type=7, attr=0xf,
range=[0x0000140000000000-0x0000141000000000) (65536MB)
[    0.000000] efi: mem54: type=7, attr=0xf,
range=[0x0000158000000000-0x0000159000000000) (65536MB)
[    0.000000] efi: mem55: type=7, attr=0xf,
range=[0x0000170000000000-0x0000171000000000) (65536MB)
[    0.000000] efi: mem56: type=11, attr=0x8000000000000001,
range=[0x0000000080000000-0x0000000090000000) (256MB)
[    0.000000] efi: mem57: type=11, attr=0x8000000000000001,
range=[0x00000000fed1c000-0x00000000fed20000) (0MB)
[    0.000000] efi: mem58: type=11, attr=0x8000000000000001,
range=[0x00000000ff000000-0x00000000ff200000) (2MB)
[    0.000000] efi: mem59: type=11, attr=0x8000000000000001,
range=[0x00003fdfe0000000-0x00003fdff4000000) (320MB)
[    0.000000] efi: mem60: type=11, attr=0x8000000000000001,
range=[0x00003fe060000000-0x00003fe074000000) (320MB)
[    0.000000] efi: mem61: type=11, attr=0x8000000000000001,
range=[0x00003fe0e0000000-0x00003fe0f4000000) (320MB)
[    0.000000] efi: mem62: type=11, attr=0x8000000000000001,
range=[0x00003fe160000000-0x00003fe174000000) (320MB)
[    0.000000] efi: mem63: type=11, attr=0x8000000000000001,
range=[0x00003fe1e0000000-0x00003fe1f4000000) (320MB)
[    0.000000] efi: mem64: type=11, attr=0x8000000000000001,
range=[0x00003fe260000000-0x00003fe274000000) (320MB)
[    0.000000] efi: mem65: type=11, attr=0x8000000000000001,
range=[0x00003fe2e0000000-0x00003fe2f4000000) (320MB)
[    0.000000] efi: mem66: type=11, attr=0x8000000000000001,
range=[0x00003fe360000000-0x00003fe374000000) (320MB)
[    0.000000] efi: mem67: type=11, attr=0x8000000000000001,
range=[0x00003fe3e0000000-0x00003fe3f4000000) (320MB)
[    0.000000] efi: mem68: type=11, attr=0x8000000000000001,
range=[0x00003fe460000000-0x00003fe474000000) (320MB)
[    0.000000] efi: mem69: type=11, attr=0x8000000000000001,
range=[0x00003fe4e0000000-0x00003fe4f4000000) (320MB)
[    0.000000] efi: mem70: type=11, attr=0x8000000000000001,
range=[0x00003fe560000000-0x00003fe574000000) (320MB)
[    0.000000] efi: mem71: type=11, attr=0x8000000000000001,
range=[0x00003fe5e0000000-0x00003fe5f4000000) (320MB)
[    0.000000] efi: mem72: type=11, attr=0x8000000000000001,
range=[0x00003fe660000000-0x00003fe674000000) (320MB)
[    0.000000] efi: mem73: type=11, attr=0x8000000000000001,
range=[0x00003fe6e0000000-0x00003fe6f4000000) (320MB)
[    0.000000] efi: mem74: type=11, attr=0x8000000000000001,
range=[0x00003fe760000000-0x00003fe774000000) (320MB)
[    0.000000] efi: mem75: type=11, attr=0x8000000000000001,
range=[0x00003fe7e0000000-0x00003fe7f4000000) (320MB)

> 
> _______________________________________________
> kexec mailing list
> kexec@lists.infradead.org
> http://lists.infradead.org/mailman/listinfo/kexec



_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

Re: [PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[Vivek Goyal]

On Fri, Dec 20, 2013 at 10:08:08AM +0900, HATAYAMA Daisuke wrote:

[..]
> 
> >cat /proc/iomem:
> >00000000-00000fff : reserved
> >00001000-0009ffff : System RAM
> >000a0000-000bffff : PCI Bus 0000:00
> >000f0000-000fffff : System ROM
> >00100000-3d162017 : System RAM
> >   01000000-015cab9b : Kernel code
> >   015cab9c-019beb3f : Kernel data
> >   01b4f000-01da9fff : Kernel bss
> >   30000000-37ffffff : Crash kernel
> >3d162018-3d171e57 : System RAM
> >3d171e58-3d172017 : System RAM
> >3d172018-3d17ae57 : System RAM
> >3d17ae58-3dc10fff : System RAM
> 
> this part is consecutive but somehow is divided into 4 entries.
> You called your environment as ``EFI virtual machine'', could you tell
> me precisely what it mean? qemu/KVM or VMware guest system? I do want
> to understand how this kind of memory map was created. I think this
> kind of memory mapping is odd and I guess this is caused by the fact
> that the system is a virtual environment.
> 
> And for Vivek, this case is a concrete example of multiple RAM entries
> appearing in a single page I suspected in the mmap failure patch,
> although these entries are consecutive in physical address and can be
> represented by a single entry by merging them in a single entry. But
> then it seems to me that there could be more odd case that multiple
> RAM entries but not consecutive. I again think this should be addressed
> in the patch for the mmap failure issue. How do you think?

Hi Hatayama,

This indeed looks very odd. See if a very small number of systems have it,
the only thing we will do is allocate extra page in second kernel for
a memory range. It will not make mmap() fail. So it is just a matter of
optimization.

Given the fact I have not seen many systems with this anomaly, I am not
too worried about it even if you don't this optimization in your patch
series. We can always take care of it later if need be.

At the same time, if you feel strongly about it and want to fix it in
same patch series, I don't mind.

Thanks
Vivek

_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

Re: [PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[HATAYAMA Daisuke]

(2013/12/20 18:00), Dave Young wrote:
> On 12/20/13 at 05:49pm, HATAYAMA Daisuke wrote:
>> (2013/12/20 11:17), Dave Young wrote:
>>>> Also, I'm interested in the memory map passed to from EFI in that
>>>>
>>>>> cat /proc/iomem:
>>>>> 00000000-00000fff : reserved
>>>>> 00001000-0009ffff : System RAM
>>>>> 000a0000-000bffff : PCI Bus 0000:00
>>>>> 000f0000-000fffff : System ROM
>>>>> 00100000-3d162017 : System RAM
>>>>>    01000000-015cab9b : Kernel code
>>>>>    015cab9c-019beb3f : Kernel data
>>>>>    01b4f000-01da9fff : Kernel bss
>>>>>    30000000-37ffffff : Crash kernel
>>>>> 3d162018-3d171e57 : System RAM
>>>>> 3d171e58-3d172017 : System RAM
>>>>> 3d172018-3d17ae57 : System RAM
>>>>> 3d17ae58-3dc10fff : System RAM
>>>>
>>>> this part is consecutive but somehow is divided into 4 entries.
>>>> You called your environment as ``EFI virtual machine'', could you tell
>>>> me precisely what it mean? qemu/KVM or VMware guest system? I do want
>>>> to understand how this kind of memory map was created. I think this
>>>> kind of memory mapping is odd and I guess this is caused by the fact
>>>> that the system is a virtual environment.
>>>
>>> This is not specific to EFI machine, it's the reserved setup_data regions
>>> They happened to be continous but they do not have to be continuous.
>>>
>>
>> Thanks for pointing out that. I've just read Documentation/x86/boot.txt
>> and parse_setup_data().
>>
>> But I don't understand well why these regions are divided as these.
>> I guess kernel divides the System RAM this way and the memory map first
>> passed to by EFI is all page aligned, right?
>
> setup_data are passed as a link list by boot loader, each node is a block
> of memory, there could be many different setup_data type.
>
>>
>> Also, looking at parse_setup_data(), currently handled data in setup_data
>> interface is extended e820 entries and dtb case only.
>>
>>                  switch (data_type) {
>>                  case SETUP_E820_EXT:
>>                          parse_e820_ext(pa_data, data_len);
>>                          break;
>>                  case SETUP_DTB:
>>                          add_dtb(pa_data);
>>                          break;
>>                  default:
>>                          break;
>>                  }
>>
>> Is it right that this kind of memory map doesn't occur as long as either
>> of information is passed to via setup_data? IOW, is this necessary
>> information?
>
> If bootloader does not pass it, there will be no such mem ranges in /proc/iomem.
>
>>
>>>>
>>>> And for Vivek, this case is a concrete example of multiple RAM entries
>>>> appearing in a single page I suspected in the mmap failure patch,
>>>> although these entries are consecutive in physical address and can be
>>>> represented by a single entry by merging them in a single entry. But
>>>> then it seems to me that there could be more odd case that multiple
>>>> RAM entries but not consecutive. I again think this should be addressed
>>>> in the patch for the mmap failure issue. How do you think?
>>>
>>> They are different problems, the previous mmap bug is for cross page regions
>>> with different page flags.
>>>
>>
>> I understand that. What I think problem here is the case where multiple
>> System RAM entries appear in a single page. In the above memory map, they
>> are 3d171000, 3d172000 and 3d17a000. My fixing patch is to copy fractional
>> pages in the 2nd kernel in order to make it possible to mmap without affecting
>> non-System RAM area as much as possible, and then if there is this kind of
>> System RAM entries, we need to use the same page in the 2nd kernel for
>> different System RAM entries that shares the same page in the 1st kernel. This
>> needs a little additional processing and we want to keep implementation as
>> simple as possible as long as there's no such system in real world. However,
>> I'm surprised to see the memory mapping above.
>
> These ranges are "system ram" of type E820_RESERVED_KERN, please see below:
> arch/x86/kernel/setup.c: e820_reserve_setup_data()
>

Thanks for explaining this. I was confused E820_RESERVED_KERN appeared in
/proc/iomem as Reserved...

-- 
Thanks.
HATAYAMA, Daisuke


_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

Re: [PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[HATAYAMA Daisuke]

(2013/12/20 21:58), Lisa Mitchell wrote:
> On Fri, 2013-12-20 at 14:13 +0000, Vivek Goyal wrote:
>> On Fri, Dec 20, 2013 at 10:08:08AM +0900, HATAYAMA Daisuke wrote:
>>
>> [..]
>>>
>>>> cat /proc/iomem:
>>>> 00000000-00000fff : reserved
>>>> 00001000-0009ffff : System RAM
>>>> 000a0000-000bffff : PCI Bus 0000:00
>>>> 000f0000-000fffff : System ROM
>>>> 00100000-3d162017 : System RAM
>>>>    01000000-015cab9b : Kernel code
>>>>    015cab9c-019beb3f : Kernel data
>>>>    01b4f000-01da9fff : Kernel bss
>>>>    30000000-37ffffff : Crash kernel
>>>> 3d162018-3d171e57 : System RAM
>>>> 3d171e58-3d172017 : System RAM
>>>> 3d172018-3d17ae57 : System RAM
>>>> 3d17ae58-3dc10fff : System RAM
>>>
>>> this part is consecutive but somehow is divided into 4 entries.
>>> You called your environment as ``EFI virtual machine'', could you tell
>>> me precisely what it mean? qemu/KVM or VMware guest system? I do want
>>> to understand how this kind of memory map was created. I think this
>>> kind of memory mapping is odd and I guess this is caused by the fact
>>> that the system is a virtual environment.
>>>
>>> And for Vivek, this case is a concrete example of multiple RAM entries
>>> appearing in a single page I suspected in the mmap failure patch,
>>> although these entries are consecutive in physical address and can be
>>> represented by a single entry by merging them in a single entry. But
>>> then it seems to me that there could be more odd case that multiple
>>> RAM entries but not consecutive. I again think this should be addressed
>>> in the patch for the mmap failure issue. How do you think?
>>
>> Hi Hatayama,
>>
>> This indeed looks very odd. See if a very small number of systems have it,
>> the only thing we will do is allocate extra page in second kernel for
>> a memory range. It will not make mmap() fail. So it is just a matter of
>> optimization.
>>
>> Given the fact I have not seen many systems with this anomaly, I am not
>> too worried about it even if you don't this optimization in your patch
>> series. We can always take care of it later if need be.
>>
>> At the same time, if you feel strongly about it and want to fix it in
>> same patch series, I don't mind.
>>
>> Thanks
>> Vivek
>
> Did I get this same segmentation fault?   It happened a few times on a
> 3.10 based kernel on a large EFI based system , but then hasn't repeated
> in further testing on this machine. This machine had no virtualization
> active. Here is partial console log of dump process:
>
> =============================================================================
>
> kdump: dump target is /dev/mapper/mpathc3
> kdump: saving to /sysroot//var/crash/127.0.0.1-2013.11.11-09:58:20/
> kdump: saving vmcore-dmesg.txt
> kdump: saving vmcore-dmesg.txt complete
> kdump: saving vmcore
> [   46.017827] scsi 20:0:0:0: Direct-Access              Geek Squad
> 8192 PQ: 0 ANSI: 0 CCS
> [   46.027849] scsi 20:0:0:0: alua: supports implicit and explicit TPGS
> [   46.035641] scsi 20:0:0:0: alua: No target port descriptors found
> [   46.042491] scsi 20:0:0:0: alua: not attached
> [   46.048152] sd 20:0:0:0: [sdak] 15663104 512-byte logical blocks:
> (8.01 GB/7.46 GiB)
> [   46.057892] sd 20:0:0:0: [sdak] Write Protect is off
> [   46.064516] sd 20:0:0:0: [sdak] No Caching mode page present
> [   46.070877] sd 20:0:0:0: [sdak] Assuming drive cache: write through
> [   46.081639] sd 20:0:0:0: [sdak] No Caching mode page present
> [   46.088035] sd 20:0:0:0: [sdak] Assuming drive cache: write through
> [   46.107070]  sdak: sdak1 sdak2 sdak3
> [   46.114636] sd 20:0:0:0: [sdak] No Caching mode page present
> [   46.121021] sd 20:0:0:0: [sdak] Assuming drive cache: write through
> [   46.128060] sd 20:0:0:0: [sdak] Attached SCSI removable disk
>
> Excluding unnecessary pages        : [  0 %]
> Excluding unnecessary
> sary pages        : [  0 %]
> Excluding unnecessary pages        : [  0
> %]
> Excluding unnecessary pages        : [  8 %]
> Excluding unnecessary
> pages        : [ 15 %]
> Excluding unnecessary pages        : [ 21 %] [
> 53.800116] usb 4-1.6: device descriptor read/64, error -110
>
> Excluding unnecessary pages        : [ 28 %]
> Excluding unnecessary
> pages        : [ 34 %]
> Excluding unnecessary pages        : [ 41 %]
> Excluding unnecessary pages        : [100 %] [   59.174433] scsi
> 18:0:0:0: Enclosure         HP       P2000 G3 FC      T240 PQ: 0 ANSI: 5
> [   59.183866] scsi 18:0:0:0: alua: supports implicit TPGS
> [   59.190190] scsi 18:0:0:0: alua: port group 01 rel port 05
> [   59.197694] scsi 18:0:0:0: alua: transition timeout set to 60 seconds
>
> ......
>
> [   76.968265] makedumpfile[1291]: segfault at 7f5ac0c39010 ip
> 00000000004297fd sp 00007fff691683a0 error 4 in makedumpfile[400000
> +46000]
> //lib/dracut/hooks/pre-pivot/9999-kdump.sh: line 88:  1291 Segmentation
> fault   [   76.996926] sd 18:0:1:51: [sdao] Synchronizing SCSI cache
>     $CORE_COLLECT[   77.003416] sd 18:0:1:50: [sdan] Synchronizing SCSI
> cache
> OR /proc/vmcore [   77.010991] sd 18:0:0:51: [sdam] Synchronizing SCSI
> cache
> $_mp/$KDUMP_PATH[   77.018590] sd 18:0:0:50: [sdal] Synchronizing SCSI
> cache
> /$HOST_IP-$DATED[   77.026611] sd 14:0:1:71: [sdaj] Synchronizing SCSI
> cache
> IR/vmcore-incomp[   77.033849] sd 14:0:1:70: [sdai] Synchronizing SCSI
> cache
> lete
> kdump: sav[   77.041467] sd 14:0:0:71: [sdah] Synchronizing SCSI cache
> ing vmcore faile[   77.049146] sd 14:0:0:70: [sdag] Synchronizing SCSI
> cached
> Rebooting.
>
>
> =======================================================================
>
> I do not have a current /proc/iomem output to go with the above.
> However this dump was nr_cpus=8 during crashkernel boot, with a 3.10
> kernel with Daisuke's version 4 of Daisuke Hatayama's patch to allow
> multi-cpu crashkernel boot.
>
> I have the efi memory map displayed as Linux booted, before this dump if
> that helps
>
> 0.000000] efi:  ACPI=0x73ffe000  ACPI 2.0=0x73ffe014  SMBIOS=0x72ef8000
> [    0.000000] efi: mem00: type=3, attr=0xf,
> range=[0x0000000000000000-0x0000000000001000) (0MB)
> [    0.000000] efi: mem01: type=2, attr=0xf,
> range=[0x0000000000001000-0x0000000000004000) (0MB)
> [    0.000000] efi: mem02: type=7, attr=0xf,
> range=[0x0000000000004000-0x000000000008e000) (0MB)
> [    0.000000] efi: mem03: type=0, attr=0xf,
> range=[0x000000000008e000-0x0000000000090000) (0MB)
> [    0.000000] efi: mem04: type=7, attr=0xf,
> range=[0x0000000000090000-0x00000000000a0000) (0MB)
> [    0.000000] efi: mem05: type=7, attr=0xf,
> range=[0x0000000000100000-0x0000000001000000) (15MB)
> [    0.000000] efi: mem06: type=2, attr=0xf,
> range=[0x0000000001000000-0x0000000002268000) (18MB)
> [    0.000000] efi: mem07: type=7, attr=0xf,
> range=[0x0000000002268000-0x0000000010000000) (221MB)
> [    0.000000] efi: mem08: type=3, attr=0xf,
> range=[0x0000000010000000-0x0000000010066000) (0MB)
> [    0.000000] efi: mem09: type=7, attr=0xf,
> range=[0x0000000010066000-0x0000000029dfb000) (413MB)
> [    0.000000] efi: mem10: type=2, attr=0xf,
> range=[0x0000000029dfb000-0x0000000039ba0000) (253MB)
> [    0.000000] efi: mem11: type=4, attr=0xf,
> range=[0x0000000039ba0000-0x000000003e179000) (69MB)
> [    0.000000] efi: mem12: type=3, attr=0xf,
> range=[0x000000003e179000-0x000000003e55c000) (3MB)
> [    0.000000] efi: mem13: type=4, attr=0xf,
> range=[0x000000003e55c000-0x000000003e567000) (0MB)
> [    0.000000] efi: mem14: type=3, attr=0xf,
> range=[0x000000003e567000-0x000000003e6e9000) (1MB)
> [    0.000000] efi: mem15: type=4, attr=0xf,
> range=[0x000000003e6e9000-0x000000003e6f4000) (0MB)
> [    0.000000] efi: mem16: type=3, attr=0xf,
> range=[0x000000003e6f4000-0x000000003e88f000) (1MB)
> [    0.000000] efi: mem17: type=4, attr=0xf,
> range=[0x000000003e88f000-0x000000003e8f4000) (0MB)
> [    0.000000] efi: mem18: type=3, attr=0xf,
> range=[0x000000003e8f4000-0x000000003e9cf000) (0MB)
> [    0.000000] efi: mem19: type=4, attr=0xf,
> range=[0x000000003e9cf000-0x000000003e9d0000) (0MB)
> [    0.000000] efi: mem20: type=3, attr=0xf,
> range=[0x000000003e9d0000-0x000000003ee9d000) (4MB)
> [    0.000000] efi: mem21: type=4, attr=0xf,
> range=[0x000000003ee9d000-0x000000003ef9b000) (0MB)
> [    0.000000] efi: mem22: type=0, attr=0xf,
> range=[0x000000003ef9b000-0x000000003efab000) (0MB)
> [    0.000000] efi: mem23: type=4, attr=0xf,
> range=[0x000000003efab000-0x000000006ba1b000) (714MB)
> [    0.000000] efi: mem24: type=10, attr=0xf,
> range=[0x000000006ba1b000-0x000000006ca1b000) (16MB)
> [    0.000000] efi: mem25: type=4, attr=0xf,
> range=[0x000000006ca1b000-0x00000000709ff000) (63MB)
> [    0.000000] efi: mem26: type=7, attr=0xf,
> range=[0x00000000709ff000-0x0000000070a10000) (0MB)
> [    0.000000] efi: mem27: type=2, attr=0xf,
> range=[0x0000000070a10000-0x0000000070c34000) (2MB)
> [    0.000000] efi: mem28: type=7, attr=0xf,
> range=[0x0000000070c34000-0x0000000070c35000) (0MB)
> [    0.000000] efi: mem29: type=2, attr=0xf,
> range=[0x0000000070c35000-0x0000000070dff000) (1MB)
> [    0.000000] efi: mem30: type=7, attr=0xf,
> range=[0x0000000070dff000-0x0000000070fa9000) (1MB)
> [    0.000000] efi: mem31: type=1, attr=0xf,
> range=[0x0000000070fa9000-0x00000000711ff000) (2MB)
> [    0.000000] efi: mem32: type=7, attr=0xf,
> range=[0x00000000711ff000-0x000000007120e000) (0MB)
> [    0.000000] efi: mem33: type=3, attr=0xf,
> range=[0x000000007120e000-0x00000000721ff000) (15MB)
> [    0.000000] efi: mem34: type=6, attr=0x800000000000000f,
> range=[0x00000000721ff000-0x00000000725ff000) (4MB)
> [    0.000000] efi: mem35: type=5, attr=0x800000000000000f,
> range=[0x00000000725ff000-0x0000000072dff000) (8MB)
> [    0.000000] efi: mem36: type=0, attr=0xf,
> range=[0x0000000072dff000-0x0000000072eff000) (1MB)
> [    0.000000] efi: mem37: type=10, attr=0xf,
> range=[0x0000000072eff000-0x0000000073eff000) (16MB)
> [    0.000000] efi: mem38: type=9, attr=0xf,
> range=[0x0000000073eff000-0x0000000073fff000) (1MB)
> [    0.000000] efi: mem39: type=4, attr=0xf,
> range=[0x0000000073fff000-0x000000007c000000) (128MB)
> [    0.000000] efi: mem40: type=7, attr=0xf,
> range=[0x0000000100000000-0x0000001080000000) (63488MB)
> [    0.000000] efi: mem41: type=7, attr=0xf,
> range=[0x0000020000000000-0x0000021000000000) (65536MB)
> [    0.000000] efi: mem42: type=7, attr=0xf,
> range=[0x0000038000000000-0x0000039000000000) (65536MB)
> [    0.000000] efi: mem43: type=7, attr=0xf,
> range=[0x0000050000000000-0x0000051000000000) (65536MB)
> [    0.000000] efi: mem44: type=7, attr=0xf,
> range=[0x0000068000000000-0x0000069000000000) (65536MB)
> [    0.000000] efi: mem45: type=7, attr=0xf,
> range=[0x0000080000000000-0x0000081000000000) (65536MB)
> [    0.000000] efi: mem46: type=7, attr=0xf,
> range=[0x0000098000000000-0x0000099000000000) (65536MB)
> [    0.000000] efi: mem47: type=7, attr=0xf,
> range=[0x00000b0000000000-0x00000b1000000000) (65536MB)
> [    0.000000] efi: mem48: type=7, attr=0xf,
> range=[0x00000c8000000000-0x00000c9000000000) (65536MB)
> [    0.000000] efi: mem49: type=7, attr=0xf,
> range=[0x00000e0000000000-0x00000e1000000000) (65536MB)
> [    0.000000] efi: mem50: type=7, attr=0xf,
> range=[0x00000f8000000000-0x00000f9000000000) (65536MB)
> [    0.000000] efi: mem51: type=7, attr=0xf,
> range=[0x0000110000000000-0x0000111000000000) (65536MB)
> [    0.000000] efi: mem52: type=7, attr=0xf,
> range=[0x0000128000000000-0x0000129000000000) (65536MB)
> [    0.000000] efi: mem53: type=7, attr=0xf,
> range=[0x0000140000000000-0x0000141000000000) (65536MB)
> [    0.000000] efi: mem54: type=7, attr=0xf,
> range=[0x0000158000000000-0x0000159000000000) (65536MB)
> [    0.000000] efi: mem55: type=7, attr=0xf,
> range=[0x0000170000000000-0x0000171000000000) (65536MB)
> [    0.000000] efi: mem56: type=11, attr=0x8000000000000001,
> range=[0x0000000080000000-0x0000000090000000) (256MB)
> [    0.000000] efi: mem57: type=11, attr=0x8000000000000001,
> range=[0x00000000fed1c000-0x00000000fed20000) (0MB)
> [    0.000000] efi: mem58: type=11, attr=0x8000000000000001,
> range=[0x00000000ff000000-0x00000000ff200000) (2MB)
> [    0.000000] efi: mem59: type=11, attr=0x8000000000000001,
> range=[0x00003fdfe0000000-0x00003fdff4000000) (320MB)
> [    0.000000] efi: mem60: type=11, attr=0x8000000000000001,
> range=[0x00003fe060000000-0x00003fe074000000) (320MB)
> [    0.000000] efi: mem61: type=11, attr=0x8000000000000001,
> range=[0x00003fe0e0000000-0x00003fe0f4000000) (320MB)
> [    0.000000] efi: mem62: type=11, attr=0x8000000000000001,
> range=[0x00003fe160000000-0x00003fe174000000) (320MB)
> [    0.000000] efi: mem63: type=11, attr=0x8000000000000001,
> range=[0x00003fe1e0000000-0x00003fe1f4000000) (320MB)
> [    0.000000] efi: mem64: type=11, attr=0x8000000000000001,
> range=[0x00003fe260000000-0x00003fe274000000) (320MB)
> [    0.000000] efi: mem65: type=11, attr=0x8000000000000001,
> range=[0x00003fe2e0000000-0x00003fe2f4000000) (320MB)
> [    0.000000] efi: mem66: type=11, attr=0x8000000000000001,
> range=[0x00003fe360000000-0x00003fe374000000) (320MB)
> [    0.000000] efi: mem67: type=11, attr=0x8000000000000001,
> range=[0x00003fe3e0000000-0x00003fe3f4000000) (320MB)
> [    0.000000] efi: mem68: type=11, attr=0x8000000000000001,
> range=[0x00003fe460000000-0x00003fe474000000) (320MB)
> [    0.000000] efi: mem69: type=11, attr=0x8000000000000001,
> range=[0x00003fe4e0000000-0x00003fe4f4000000) (320MB)
> [    0.000000] efi: mem70: type=11, attr=0x8000000000000001,
> range=[0x00003fe560000000-0x00003fe574000000) (320MB)
> [    0.000000] efi: mem71: type=11, attr=0x8000000000000001,
> range=[0x00003fe5e0000000-0x00003fe5f4000000) (320MB)
> [    0.000000] efi: mem72: type=11, attr=0x8000000000000001,
> range=[0x00003fe660000000-0x00003fe674000000) (320MB)
> [    0.000000] efi: mem73: type=11, attr=0x8000000000000001,
> range=[0x00003fe6e0000000-0x00003fe6f4000000) (320MB)
> [    0.000000] efi: mem74: type=11, attr=0x8000000000000001,
> range=[0x00003fe760000000-0x00003fe774000000) (320MB)
> [    0.000000] efi: mem75: type=11, attr=0x8000000000000001,
> range=[0x00003fe7e0000000-0x00003fe7f4000000) (320MB)
>

This is an original memory map passed to by UEFI. kexec tool
uses /proc/iomem, which can be modified during boot by kernel.
Although the above EFI memory map is all page-size aligned,
the one in /proc/iomem might be not so.

The best way to check the bug here is to see a full back trace
of makedumpfile to see where and how it resulted in
segmentation fault, just as Wang did.

If you feel awkward to get process core dump of makedumpfile
in the 2nd kernel, it's simpler to first get crash dump by
cp command to keep shape of /proc/vmcore and then use makedumpfile
to the generated crash dump. If makedumpfile crashes in the 2nd
kernel, it also crashes in the 1st kernel, since the crash dump
collected by cp command is the same as /proc/vmcore.

-- 
Thanks.
HATAYAMA, Daisuke


_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec

Re: [PATCH] makedumpfile: memset() in cyclic bitmap initialization introduce segment fault

[HATAYAMA Daisuke]

(2013/12/20 23:13), Vivek Goyal wrote:
> On Fri, Dec 20, 2013 at 10:08:08AM +0900, HATAYAMA Daisuke wrote:
>
> [..]
>>
>>> cat /proc/iomem:
>>> 00000000-00000fff : reserved
>>> 00001000-0009ffff : System RAM
>>> 000a0000-000bffff : PCI Bus 0000:00
>>> 000f0000-000fffff : System ROM
>>> 00100000-3d162017 : System RAM
>>>    01000000-015cab9b : Kernel code
>>>    015cab9c-019beb3f : Kernel data
>>>    01b4f000-01da9fff : Kernel bss
>>>    30000000-37ffffff : Crash kernel
>>> 3d162018-3d171e57 : System RAM
>>> 3d171e58-3d172017 : System RAM
>>> 3d172018-3d17ae57 : System RAM
>>> 3d17ae58-3dc10fff : System RAM
>>
>> this part is consecutive but somehow is divided into 4 entries.
>> You called your environment as ``EFI virtual machine'', could you tell
>> me precisely what it mean? qemu/KVM or VMware guest system? I do want
>> to understand how this kind of memory map was created. I think this
>> kind of memory mapping is odd and I guess this is caused by the fact
>> that the system is a virtual environment.
>>
>> And for Vivek, this case is a concrete example of multiple RAM entries
>> appearing in a single page I suspected in the mmap failure patch,
>> although these entries are consecutive in physical address and can be
>> represented by a single entry by merging them in a single entry. But
>> then it seems to me that there could be more odd case that multiple
>> RAM entries but not consecutive. I again think this should be addressed
>> in the patch for the mmap failure issue. How do you think?
>
> Hi Hatayama,
>
> This indeed looks very odd. See if a very small number of systems have it,
> the only thing we will do is allocate extra page in second kernel for
> a memory range. It will not make mmap() fail. So it is just a matter of
> optimization.
>

Yes, mmap doesn't fail. Without the optimization, we get the first System RAM
data only from multiple System RAM entries in a single page. vmcore_list
contains entries for each entry in the multiple System RAM entries, although
we cannot look up the entries except for the 1st one since they have the same
offset.

> Given the fact I have not seen many systems with this anomaly, I am not
> too worried about it even if you don't this optimization in your patch
> series. We can always take care of it later if need be.
>
> At the same time, if you feel strongly about it and want to fix it in
> same patch series, I don't mind.
>

I think some part of System RAM is dropped off from crash dump is a problem.
But I also think it important to fix this issue as soon as possible. So,
I want to introduce basic copying mechanism first, and then focus on the
optimization.

By the way, I guess one of what you worry about is how to make sure whether
the logic of dividing each System RAM area into at most three parts, is
correct or not. Is it better to describe a simple proof somewhere as comment?

-- 
Thanks.
HATAYAMA, Daisuke


_______________________________________________
kexec mailing list
kexec@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/kexec