I bet this is due to:

        vaddr = ioremap_cache(pfn << PAGE_SHIFT, PAGE_SIZE);
        if (!vaddr)
                return -ENOMEM;

It is really confusing that it returns ENOMEM for read(2)...

Did you see any kernel messages? If not, try to add "ignore_loglevel" to the second kernel.

Might be related with:

commit 23da4563659382c6da6ac1fdf70099fe12010894
Author: Neil Horman <nhorman>
Date:   Thu Oct 14 18:54:30 2010 -0400

    [kdump] kexec: accelerate vmcore copies by marking oldmem in /proc/vmcore as cached

which is merged in kernel-2.6.32-85.el6~35, so does kernel-2.6.32-84.el6 have this problem on that machine?

readmem: type_addr: 0, addr:ffff880000040d40, size:4 

BIOS-provided physical RAM map: 
 BIOS-e820: 0000000000000100 - 00000000000a0000 (usable) 
 BIOS-e820: 0000000000100000 - 00000000cffa8000 (usable) 
...
user-defined physical RAM map: 
 user: 0000000000000000 - 0000000000001000 (reserved) 
 user: 0000000000001000 - 00000000000a0000 (usable) 
...
Seems 640K related...

*** Bug 690362 has been marked as a duplicate of this bug. ***

Tested on amd-dinar-03.lab.bos.redhat.com with latest build:
======================================================================
[root@amd-dinar-03 ~]# rpm -q kernel kexec-tools
kernel-2.6.32-131.0.1.el6.x86_64
kexec-tools-2.0.0-186.el6.x86_64
[root@amd-dinar-03 ~]# tail /etc/kdump.conf 
#kdump_post /var/crash/scripts/kdump-post.sh
#extra_bins /usr/bin/lftp
#disk_timeout 30
#extra_modules gfs2
#options modulename options
#default shell

ext4 /dev/mapper/vg_amddinar03-lv_root
core_collector makedumpfile -c --message-level 1 -d 31
default shell
[root@amd-dinar-03 ~]# service kdump restart
Stopping kdump:[  OK  ]
Detected change(s) the following file(s):
  
  /etc/kdump.conf
Rebuilding /boot/initrd-2.6.32-131.0.1.el6.x86_64kdump.img
Your running kernel is using more than 70% of the amount of space you reserved for kdump, you should consider increasing your crashkernel reservation[WARNING]
Starting kdump:[  OK  ]
[root@amd-dinar-03 ~]# insmod /mnt/tests/kernel/kdump/crash-crasher/crasher/crasher.ko
[root@amd-dinar-03 ~]# echo 0 > /proc/crasher
--------------------------------------------------------------------

Free memory/Total memory (free %): 187912 / 241096 ( 77.9407 )
Scanning logical volumes
  Reading all physical volumes.  This may take a while...
  Found volume group "vg_amddinar03" using metadata type lvm2
Activating logical volumes
  3 logical volume(s) in volume group "vg_amddinar03" now active
Free memory/Total memory (free %): 186948 / 241096 ( 77.5409 )
Saving to the local filesystem /dev/mapper/vg_amddinar03-lv_root
e2fsck 1.41.12 (17-May-2010)
/dev/mapper/vg_amddinar03-lv_root: recovering journal
Clearing orphaned inode 2895036 (uid=0, gid=0, mode=0100600, size=4096)
Clearing orphaned inode 2895000 (uid=0, gid=0, mode=0100600, size=4096)
Clearing orphaned inode 2894999 (uid=0, gid=0, mode=0100600, size=4096)
Clearing orphaned inode 2894990 (uid=0, gid=0, mode=0100600, size=4096)
/dev/mapper/vg_amddinar03-lv_root: clean, 122390/3276800 files, 8423377/13107200 blocks
EXT4-fs (dm-0): mounted filesystem with ordered data mode
Free memory/Total memory (free %): 185768 / 241096 ( 77.0515 )
Loading SELINUX policy
type=1404 audit(1303351095.630:2): enforcing=1 old_enforcing=0 auid=4294967295 ses=4294967295
type=1403 audit(1303351096.448:3): policy loaded auid=4294967295 ses=4294967295
Copying data                       : [100 %] 
Saving core complete
md: stopping all md devices.
sd 0:0:1:0: [sda] Synchronizing SCSI cache
Restarting system.
machine restart


Change status to VERIFIED

An advisory has been issued which should help the problem
described in this bug report. This report is therefore being
closed with a resolution of ERRATA. For more information
on therefore solution and/or where to find the updated files,
please follow the link below. You may reopen this bug report
if the solution does not work for you.

http://rhn.redhat.com/errata/RHBA-2011-0736.html