Sponsored Content
Operating Systems AIX AIX 6.1, POWER5 and Spectre/Meltdown Post 303012009 by MichaelFelt on Saturday 27th of January 2018 02:46:11 PM
Old 01-27-2018
Quote:
Originally Posted by richardsantink
AIX 7.2 will only install on Power7 and newer.

Meltdown and Spectre affect all Power systems:
IBM Systems Magazine - Security Vulnerability Impacts POWER Processors
While I am not an expert in this area - I do recall POWER6 does not do branch prediction. It is blazes ahead - and if the instruction path is wrong - then the pre-fetch is just thrown away.

The idea was that the tremendous jump in clock-speed was enough that the 'occasional' missed prediction was worth it. In other words - processor heat was from raw speed, rather than from parallel calculations computing branch prediction.

Again, not an expert - whether these vulnerabilities stem from any pre-fetch, or only from "predicted branch pre-fetch" - I don't know.

Yes, it's out there - and I am actually a bit more curious to hear about the 'monitors' or signature-scanners that get built to spot anything attempting to exploit it.
 

7 More Discussions You Might Find Interesting

1. Filesystems, Disks and Memory

hard disk meltdown

I had an issue with a second hard disk in my machine. I have a sparc station running solaris 7. It was working fine but now it wont mount on boot up and when you try to mount it manually it gives an I/O error. I tried a different disk as a control which was fine. What I want to know is if my... (3 Replies)
Discussion started by: Henrik
3 Replies

2. AIX

Problem with knlist in AIX 5.3 for Power5 processor

Hi all, I have a AIX 5.3 Installation on the Power5 processor. I am using the knlist API for getting the details of the CPU details like user cpu, kernel cpu, idle cpu and wait cpu. My code is working fine on other Installations of AIX like PowerPC_POWER4 but it is giving wrong values on Power5... (0 Replies)
Discussion started by: rajesht
0 Replies

3. AIX

Upgrading Power5 from 5.3 to...

I have a power5 that I would like to update and bring everything forward. I am wondering about trouble-free upgrades. Is it better to go 5.3 -> 7 or 5.3 ->6 ->7? The power5 box is behind and hopefully can learn from experiences here. Thanks. (6 Replies)
Discussion started by: redi
6 Replies

4. AIX

POWER5+ 2.1Gz cpu wattage ?

Hello, What is wattage of POWER5+ 2.1Gz CPU ? It's dual-core 90nm chip. At least approximately ? thanks, Vilius (1 Reply)
Discussion started by: vilius
1 Replies

5. AIX

Power5 9111-520 reload of AIX 6.1 - help needed

Hello all, I am hoping someone can point me in the right direction here. We had a machine running AIX 6.1 that someone completely corrupted, so we decided to re-load it from media. I was able to boot into DVD media, select options for complete (destructive) install, AIX standard, etc. The... (0 Replies)
Discussion started by: JamesNJ
0 Replies

6. War Stories

Linus Torvalds reply about Meltdown and Spectre.

Apologies if this is the wrong forum but... ...This is hard hitting stuff. LKML: Linus Torvalds: Re: x86/enter: Create macros to restrict/unrestrict Indirect Branch Speculation (5 Replies)
Discussion started by: wisecracker
5 Replies

7. What is on Your Mind?

Meltdown and Spectre CPU bugs

This seems a bit serious: Meltdown, Spectre: The password theft bugs at the heart of Intel CPUs • The Register Vulnerability Note VU#584653 - CPU hardware vulnerable to side-channel attacks Project Zero: Reading privileged memory with a side-channel (8 Replies)
Discussion started by: Scrutinizer
8 Replies
emulate_branch(3)					     Library Functions Manual						 emulate_branch(3)

Name
       emulate_branch, execute_branch - branch emulation

Syntax
       #include <signal.h>

       emulate_branch(scp, branch_instruction)
       struct sigcontext *scp;
       unsigned long branch_instruction;

       execute_branch(branch_instruction)
       unsigned long branch_instruction;

Description
       The  function  is  passed  a signal context structure and a branch instruction.	It emulates the branch based on the register values in the
       signal context structure.  It modifies the value of the program counter in the signal context  structure  (sc_pc)  to  the  target  of  the
       branch_instruction.   The program counter must initially be pointing at the branch and the register values must be those at the time of the
       branch.	If the branch is not taken the program counter is advanced to point to the instruction after the delay slot (sc_pc += 8).

       If the branch instruction is a `branch on coprocessor 2' or `branch on coprocessor 3' instruction, calls to  execute  the  branch  in  data
       space to determine if it is taken or not.

Return Values
       The  function returns a 0 if the branch was emulated successfully.  A non-zero value indicates the value passed as a branch instruction was
       not a branch instruction.

       The function returns non-zero on taken branches and zero on non-taken branches.

Restrictions
       Since is only intended to be used by it does not check its parameter to see if in fact it is a branch instruction.  It is really a stop gap
       in case a coprocessor is added without the kernel fully supporting it (which is unlikely).

See Also
       cacheflush(2), sigvec(2), signal(3)

								       RISC							 emulate_branch(3)
All times are GMT -4. The time now is 09:45 PM.
Unix & Linux Forums Content Copyright 1993-2022. All Rights Reserved.
Privacy Policy