Hello Keith and all kdb
users,


                         I have to report the following bug :

I use an official linux-2.6.15 with kdb-v4.4 on IA-64 machine.
When trying to disassemble the "mpt_get_msg_frame" function (located in
drivers/message/fusion/mptbase.c),
I obtain a incoherent content on my screen :
__________________________________________________________________
[1]kdb> id mpt_get_msg_frame
0xa00000020303f7a0 fusion_exit+0xc40[MMB]       data8 0x010181e00
0xa00000020303f7a6 fusion_exit+0xc46            chk.a.nc
f0,0xa0000002033378e0
0xa00000020303f7ac fusion_exit+0xc4c            data8 0x14000000406

0xa00000020303f7b0 fusion_exit+0xc50[MMB]       data8 0x010181e00
0xa00000020303f7b6 fusion_exit+0xc56            chk.a.nc
f0,0xa0000002035978f0
0xa00000020303f7bc fusion_exit+0xc5c            data8 0x14000000406

0xa00000020303f7c0 fusion_exit+0xc60[MMB]       data8 0x010181e00
0xa00000020303f7c6 fusion_exit+0xc66            chk.a.clr
f0,0xa0000002035c3900
0xa00000020303f7cc fusion_exit+0xc6c            data8 0x14000000406

0xa00000020303f7d0 fusion_exit+0xc70[MMB]       data8 0x010181e00
0xa00000020303f7d6 fusion_exit+0xc76            chk.a.nc
f0,0xa000000203e2b910
0xa00000020303f7dc fusion_exit+0xc7c            data8 0x14000000406

0xa00000020303f7e0 fusion_exit+0xc80[MMB]       data8 0x010181e00
0xa00000020303f7e6 fusion_exit+0xc86            chk.a.clr
f0,0xa00000020360b920
0xa00000020303f7ec fusion_exit+0xc8c            data8 0x14000000406

0xa00000020303f7f0 fusion_exit+0xc90[MMB]       data8 0x010181e00
0xa00000020303f7f6 fusion_exit+0xc96            break.m 0xa00
__________________________________________________________________

I think that this pointer contains the function descriptor and not the
function address,
because when publishing the content of this area, I have :
__________________________________________________________________
[1]kdb> md8 mpt_get_msg_frame 1
0xa00000020303f728 a000000203031e00 a00000020303c018   ................
__________________________________________________________________

I first thought that it was an incoherency between module management
and kdb module address retrieving, but for some other functions,
I do not have this problem. For example, with "mptscsih_proc_info" function:
__________________________________________________________________
[1]kdb> id mptscsih_proc_info
0xa0000002030618c0 mptscsih_proc_info[MMI]       alloc r41=ar.pfs,15,11,0
0xa0000002030618c6 mptscsih_proc_info+0x6            adds r12=-32,r12
0xa0000002030618cc mptscsih_proc_info+0xc            addl r44=16408,r1

0xa0000002030618d0 mptscsih_proc_info+0x10[MMI]       adds r32=1160,r32
0xa0000002030618d6 mptscsih_proc_info+0x16            mov r42=r1
0xa0000002030618dc mptscsih_proc_info+0x1c            cmp.eq p8,p9=0,r34;;

0xa0000002030618e0 mptscsih_proc_info+0x20[MMI]       adds r38=28,r12
0xa0000002030618e6 mptscsih_proc_info+0x26            adds r14=16,r12
0xa0000002030618ec mptscsih_proc_info+0x2c            adds r39=32,r12

0xa0000002030618f0 mptscsih_proc_info+0x30[MMI]       add r44=r0,r44
0xa0000002030618f6 mptscsih_proc_info+0x36            ld8 r32=[r32]
0xa0000002030618fc mptscsih_proc_info+0x3c            cmp4.eq p6,p7=0,r37;;

0xa000000203061900 mptscsih_proc_info+0x40[MMI]       nop.m 0x0
0xa000000203061906 mptscsih_proc_info+0x46            mov r43=r14
0xa00000020306190c mptscsih_proc_info+0x4c            mov r40=b0

0xa000000203061910 mptscsih_proc_info+0x50[MMB]       nop.m 0x0
0xa000000203061916 mptscsih_proc_info+0x56            mov r8=r0
[1]more> q
[1]kdb> md8 mptscsih_proc_info 1
0xa0000002030618c0 00c00580163d4808 9200098580467e33   .H=.....3~F.....
[1]kdb>
__________________________________________________________________

Could someone tell me what it is due to ? Thanks in advance for your help,
best regards,

                                                                   François
---------------------------
Use http://oss.sgi.com/ecartis to modify your settings or to unsubscribe.

Reply via email to