1. Feature Name: enable deep C states for Nehalem Because RH is getting some late pressure from OEMs and customers for TylersburgEP/Nehalem support in 4.7, I'm concerned that this might be required. We had intended only to submit it for 5.3, but I now plan to look at it ASAP in the 4.7 context. This patch that went upstream in June 2007 gives a a 16x improvement in efficiency at idle on Nehalem chips. http://git.kernel.org/gitweb.cgi?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=18eab8550397f1f3d4b8b2c5257c88dae25d58ed 2. Description: Architectures (mark all that apply) x 32-bit x86 64-bit Itanium2 x 64-bit Intel EM64T Dependencies: External links: Priority (H,M,L): H Target Releases: Target Release Date: Drivers or hardware dependency: Target Kernel: 3. Business Justification: a. Why is this feature needed? Big power efficiency improvement b. What hardware does this enable? Nehalem CPUs c. Forecast, volume or high end platform? Volume d. Any configuration info? e. When needed? f. Are there other dependencies (drivers). 4. Status: a. Is code accepted upstream? If so, what version or patch? See above. 5. Hardware to Red Hat? Red Hat can't integrate/test features without hardware. Yes, two alpha Tylersburg EP and one Tylersburg HEDT are already at RH, more coming. 6. Partner management contact, email, phone, chat keve.a.gabbert 7. Partner technical contact, email, phone, chat jvillalo, grgustaf
The Nehalem processor is integral to the Tylersburg-EP platform that all of the OEMs will be planning. Moreover, the deep c-states improves the overall power efficiency of idle CPUs, which many customers are very interested in.
This request is not going to be addressed in RHEL 4.7. Moving to 4.8.
John, is 443516 dup of this BZ?
John, never mind! 443516 is the 5.3 equivalent.
It has been decided that this bug will be to intrusive to fix for RHEL 4.x. We will only support deep c-states in RHEL 5.3 and beyond.