The Linux-AMD AGP bug – who's to blame?

Cache, check or card?


After arranging a tete a tete between two representatives of the Linux kernel team and AMD, Gentoo founder Daniel Robbins has drawn some initial conclusions on the bug affecting Linux users on AMD Athlon AGP systems.

Daniel concludes that neither side is guilty; that it isn't an AMD bug, more a feature; but he also recommends that kernel hackers find a new problem to this particular situation. It appears that the GART (Graphics Address Remapping Table) which feeds the AGP card with system memory isn't cache coherent. Although both Linux (and apparently Windows 2000) expect it to be.

That's an drastic oversimplification of Daniel's postings on the Gentoo front page this morning
and his explanation on the kernel mailing list.

The GART lays out memory for the AGP card in a contiguous block, but the real memory that's being addressed. of course, lies all over the place - having been paged out in 4k blocks from main system memory. Intel added 4MB pages, and the temporary workaround for both Linux and Windows 2000 is to disable the 4MB page option.

AMD concludes:- "Our conclusion is that the operating system is creating coherency problems within the system by creating cacheable translation to AGP GART-mapped physical memory...
When the cache-line eviction occurs the stale data written to physical memory has fatal side effects."

With our limited knowledge of PC hardware architecture - and we trust Register readers can explain this one for us - we can't quite see how that relates to the 4k/4MB page size option. Why can't a simple flush clear the cache, we wonder? Let us know.

Nevertheless disabling 4MB pages appears to do the trick, all agree.

There's a birds eye view of where the GART fits in to the scheme of things at Anandtechhere

Related Story

AMD chip bug snares Linux users


Keep Reading

Tech Resources

What WAF is right for you

Applications are architected in many ways, but all need protection from threats. Learn the most important things to consider when choosing a WAF.

Three reasons you need a hybrid multicloud

Businesses need their IT teams to operate applications and data in a hybrid environment spanning on-premises private and public clouds. But this poses many challenges, such as managing complex networking, re-architecting applications for the cloud, and managing multiple infrastructure silos. There is a pressing need for a single platform that addresses these challenges - a hybrid multicloud built for the digital innovation era. Just this Regcast to find out: Why hybrid multicloud is the ideal path to accelerate cloud migration.

Top 20 Private Cloud Questions Answered

Download this asset for straight answers to your top private cloud questions.

How backup modernization changes the ransomware game

If the thrill of backing up your data and wondering if you will ever see it again has worn off, start the new year by getting rid of the lingering pain of legacy backup. Bipul Sinha, CEO of the Cloud Data Management Company, Rubrik, and Miguel Zatarain, Director of Global Infrastructure Technology at PACCAR, Fortune 500 manufacturer of trucks and Rubrik customer, are talking to the Reg’s Tim Phillips about how to eliminate the costly, slow and spotty performance of legacy backup, and how to modernize your implementation in 2021 to make your business more resilient.

Biting the hand that feeds IT © 1998–2021