Talk:Advanced Host Controller Interface

Why is AHCI problem? You just have to follow a specific install procedure...

...that is BEFORE you install Windows, choose what you want: AHCI or IDE. If you want AHCI then prepare the floppy disk that came with the motherboard (or make it with the motherboard CD) OR slipstream it onto Windows installation disk. Then install Windows normally. The only real danger here is if you install Windows in AHCI mode and later change it to IDE and vice versa. That causes BSODs and problems. Ivan Ivanković 21:22, 11 April 2007 (UTC)

no real 'information'

This article doesn't really talk about what AHCI is, and how it is different from the legacy compatible mode. Added to that, the same information (same word-to-word information) is available in another page http://en.wikipedia.org/wiki/AHCI). Probably, the article can continue to exist in the other page (AHCI is more popular name, than its full acronym), and this page can be deleted. —The preceding unsigned comment was added by 12.7.85.90 (talk • contribs) 21:24, January 24, 2007 (UTC)

-> The dupe is gone now. I concur fully about the lack of information. I've moved this to the top since it's the most important issue with this atricle atm. --62.231.60.207 02:13, 27 February 2007 (UTC)

See Intel AHCI Specification for details on AHCI. ( www.intel.com/technology/serialata/ahci.htm )

What is the advantage if AHCI mode was used in SATA?

Can anybody help and answer my queries? thanks —The preceding unsigned comment was added by 210.23.216.194 (talk) 06:03, 12 December 2005 (UTC)

Try the reference desk. Neutrality^talk 22:13, 4 February 2006 (UTC)

-> This article should say! We really need someone who can explain + reference the advantages. --62.231.60.207 02:13, 27 February 2007 (UTC)

-> There are several advantages to AHCI. First, one can access more than 4 drives concurrently per controller. Second, there is support for hot-plugging (or 'surprise removal' in Intel-speak). There is also support for staggered spin-up and enclosure management, though Intel and AMD implementations do not support these features at this time. Third, the programming interface is high-level. Intermediate FISes are handled in hardware. Forth, after PCI-level configuration, the programming interface is vendor independent. -- Quanstro 02:22, 13 September 2007 (UTC)

Warning relevant?

Half of the article is a warning about Windows's blue screening when the BIOS is changed to use AHCI. Is this actually appropriate for a general purpose Encyclopedia? 67.166.242.232 04:32, 26 July 2006 (UTC)

This article helped me a lot. I think it is needed especially that sometimes some mainboards change the bios setting from ata to ahci by themself only, and then the Windows XP crashes on the boot.--Doxent 11:43, 2 October 2006 (UTC)

That's nice, but it doesn't mean the information belongs here instead of a Windows help site. 67.166.242.232 12:07, 22 October 2006 (UTC)

-> I disagree: if this is a 'noteable' problem, then it should be mentioned. As per the above though - it's more important that more details be added than we argue over what's already there. I vote the we leave that section as-is. --62.231.60.207 02:13, 27 February 2007 (UTC)

--> I concur. This is a 'notable' problem. I had never heard of AHCI until it blew up a reinstallation of Windows XP over Windows Vista. Luckily for me, the bios supported a compatibility setting. I was able to install the OS in compatibility mode, install the Intel storage matrix drivers and then re-enable AHCI in the bios. The information in this article is very relevant to the problem I encountered. —Preceding unsigned comment added by 69.145.86.101 (talk) 20:06, 27 December 2007 (UTC)

add Linux support section?

AHCI is fully supported in Linux. Kernel 2.6.18 has limited support, but 2.6.19 (due out soon) will have full support. Some distributions (such as Fedora Core) already include the 2.6.19 patches in 2.6.18. —The preceding unsigned comment was added by 65.67.98.193 (talk • contribs) 15:57, November 27, 2006 (UTC)

I use fc6 and can confirm I have installed on a linux software raid across 3 disks in AHCI mode. Certainly leaves windows in it's wake... —The preceding unsigned comment was added by 203.206.56.233 (talk • contribs) 05:18, December 29, 2006 (UTC)

-> Added. Renamed the "Limitations" section to reflect it's contents. --62.231.60.207 02:13, 27 February 2007 (UTC)

neutrality

linking to bsod is questionable at least. and no, i'm not a microsoft guy —The preceding unsigned comment was added by 85.182.127.134 (talk) 15:26, 14 March 2007 (UTC)

as a computer technician i have seen the bsod problems related to ahci. —The preceding unsigned comment was added by 195.72.176.180 (talk • contribs) 12:13, March 24, 2007 (UTC)

TCT: Misconfiguration leads to BSOD when using AHCI with Windows. When the proper MSD drivers are loaded, you won't get any blue screens. —The preceding unsigned comment was added by 124.82.34.10 (talk • contribs) 10:43, April 7, 2007 (UTC)

It is true that BSoDs are caused by enabling AHCI in BIOS without first installing the drivers in Windows. What is the neutrality question? How does citing a BSoD relate to neutrality? I'm removing the neutrality notice. Justin Force 15:31, 1 May 2007 (UTC)

MSI problems on AMD/ATI chipsets

Someone mentioned that MSI needs to be disabled on certain AMD/ATI chipsets under Linux, reverting the PCIe to work as a faster PCI bus. According to the MSI article, they exist in the PCI 2.2 standard, so MSI is valid on Parallel PCI as well. What difference does MSI make? Can it be further explained in the MSI article? (I'm asking here because it's the only place I see a mention of it) 75.161.144.45 13:16, 19 July 2007 (UTC)

MSIs fix some problems wrt. the asynchronous nature of interrupt lines (and their scarcity on x86). When a device raises an IRQ line after writing data to RAM, sometimes the IRQ triggers BEFORE the previous write has finished, which requires workarounds in driver IRQ handlers. An MSI is another PCI write op to a magic address with ordering intact; so when the MSI triggers the previous write has finished an hit its destination. Furthermore, the MSI PCI write can be done using different magic values to trigger different IRQ handlers; so instead of checking an IRQ status register when the handler is triggered you use the message the device sent you with the MSI to determine irq cause (e.g. register one handler per MSI message). —Preceding unsigned comment added by 212.86.197.209 (talk) 14:05, 1 October 2007 (UTC)

Neutrality 2

This article is way too Windows and Linux centric. Information concerning AHCI and other operating systems must be added for fairness. HOWEVER, most of that information is not relevent to AHCI itself anyway. I hesitate to prune a ton of information without discussing first. --Afed 20:28, 4 December 2007 (UTC)

I think you're confusing neutrality with a lack of content. Yes, the information here is mainly limited to Windows and Linux OS,s but that alone doesn't make the article non-neutral; it just means that we need to add more information regarding AHCI and other OSs. I replaced the npov article with the missing info template, as I think this one better reflects your concerns. — EagleOne\^Talk 22:45, 4 December 2007 (UTC)

AHCI and XP

"Support for AHCI is also featured in Windows XP Media Center Edition (the only version of XP to support it), prompting many manufacturers to ship with this version despite the hardware not supporting the multimedia features of this version (an example being the Dell Dimension 5150C)."

This is not correct. While Dell machines do come with discs which have Intel's AHCI driver bundled, the disc is modified by Dell for this purpose. Generic MCE discs do not include AHCI support. Karsini (talk) 21:24, 5 December 2007 (UTC)