Skip to contentSkip to conversation

I/O APIC: Timer through 8259A revamp

DISCUSSION yuhong 2 participants 4 messages Apr 23, 2009 – Oct 3, 2009 3 messages not in the archive

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

yuhong

author
#3 lore raw
Quoted 2 lines

 The new burst of breakage came with the invention of ACPI and its tables for interrupt routing for the APIC.

Yep, when NVIDIA and ATI entered the chipset market, who would have predicted that it would cause more 8254 and APIC related troubles? Or it would be years before it were finally fixed properly? Even worse, ATI made chipsets before AMD and Intel introduced their 64-bit CPUs, yet it was the x86-64 architecture where the attempts to fix the ATI APIC 8254 timer issues were going to. Only later when it was finally fixed properly was it backported to x86.

#4 lore raw
Quoted 9 lines · yuhong
Quoted 2 lines

 The new burst of breakage came with the invention of ACPI and its tables for interrupt routing for the APIC.

Yep, when NVIDIA and ATI entered the chipset market, who would have predicted that it would cause more 8254 and APIC related troubles? Or it would be years before it were finally fixed properly? Even worse, ATI made chipsets before AMD and Intel introduced their 64-bit CPUs, yet it was the x86-64 architecture where the attempts to fix the ATI APIC 8254 timer issues were going to. Only later when it was finally fixed properly was it backported to x86.

I'm not sure what you mean here. The ATI timer fixes were done on both i386 and x86-64.

-Andi

Signature · 1 line
[email protected] -- Speaking for myself only.
author
#6 lore raw
In reply to Andi Kleen

<[email protected]>
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0

Quoted 2 lines · On Thu=2C Apr 23=2C 2009 at 10:45:03AM -0700=2C yuhong wrote:
Quoted 1 line
Quoted 1 line

The new burst of breakage came with the invention of ACPI and its tabl=

es

Quoted 11 lines
Quoted 8 lines
Quoted 1 line

for interrupt routing for the APIC.

Yep=2C when NVIDIA and ATI entered the chipset market=2C who would have predicted that it would cause more 8254 and APIC related troubles? Or it would be years before it were finally fixed properly? Even worse=2C ATI made chipsets before AMD and Intel introduced their 64-bit CPUs=2C yet it was the x86-64 architecture where the attempts to fix the ATI APIC 8254 timer issues were going to. Only later when it was finally fixed properly was it backported to x86.

I'm not sure what you mean here. The ATI timer fixes were done on both i386 and x86-64.

But why did it took so long to finally fix this when ATI has been making ch= ipsets since 2001?
Even worse=2C why was the attempts to fix this issue going to the x86-64 br= anch when ATI was making chipsets before the AMD Athlon 64 was even release= d?

Yuhong Bao
=0A=
_________________________________________________________________=0A=
Hotmail: Trusted email with powerful SPAM protection.=0A=
http://clk.atdmt.com/GBL/go/177141665/direct/01/=