Bug 136018
Summary: | Cannot run FC3T3 with included kernel on Virtual PC 2004 | ||||||
---|---|---|---|---|---|---|---|
Product: | [Fedora] Fedora | Reporter: | David Sanders <david> | ||||
Component: | kernel | Assignee: | Dave Jones <davej> | ||||
Status: | CLOSED NEXTRELEASE | QA Contact: | |||||
Severity: | high | Docs Contact: | |||||
Priority: | medium | ||||||
Version: | 3 | CC: | alex_extreme2, bamadogg, barryn, bh, brian_myers, bugz, fedorabugs, koyeung, pfrields, vbdigdaddy, wtogami | ||||
Target Milestone: | --- | ||||||
Target Release: | --- | ||||||
Hardware: | i686 | ||||||
OS: | Linux | ||||||
Whiteboard: | |||||||
Fixed In Version: | Doc Type: | Bug Fix | |||||
Doc Text: | Story Points: | --- | |||||
Clone Of: | Environment: | ||||||
Last Closed: | 2004-11-27 22:12:29 UTC | Type: | --- | ||||
Regression: | --- | Mount Type: | --- | ||||
Documentation: | --- | CRM: | |||||
Verified Versions: | Category: | --- | |||||
oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
Cloudforms Team: | --- | Target Upstream Version: | |||||
Embargoed: | |||||||
Attachments: |
|
Description
David Sanders
2004-10-16 20:22:35 UTC
Created attachment 105738 [details]
Screenshot of Fedora Core 3 Test 3 running under Virtual PC 2004
Having experimented with this previously in FC3Test1, it appears that Virtual PC does not deal well with the 4G/4G kernel (e.g., CONFIG_X86_4G = Y). Could we have an alternate kernel package with this option turned off? The .i586 kernel doesn't have 4G/4G built in (while .i686 does). I have confirmed that the 4G/4G patch is the problem. A kernel that is identical in every way except having 4G/4G mode disabled works. I will have instructions and a kernel for Fedora Core 3 when it is released. For the future, perhaps we could have a kernel-vpc package that doesn't have 4g4g and installs by default if inside of a virtual machine. I'm not sure why we need a 4g4g kernel as default anyway. I would suggest that if the amount of memory is less than 1G, we not install a 4g4g kernel anyway. The documentation says there is a 20% performace hit by using it. What do you think? which documentation states that ? The hit is much less than you think. SELinux for example probably adds more overhead. > which documentation states that? [barryn@localhost redhat]$ rpmbuild -bp --target i686 SPECS/kernel-2.6.spec [ snip... ] [barryn@localhost redhat]$ cd BUILD/kernel-2.6.9/linux-2.6.9/ [barryn@localhost linux-2.6.9]$ make menuconfig [ snip... ] Processor type and features ---> <Select> [*] 4 GB kernel-space and 4 GB user-space virtual memory support <Help> grep shows that the documentation there is in arch/i386/Kconfig. Here's the relevant quote -- pay particular attention to the last sentence: >The cost of this option is additional TLB flushes done at >system-entry points that transition from user-mode into kernel-mode. >I.e. system calls and page faults, and IRQs that interrupt user-mode >code. There's also additional overhead to kernel operations that copy >memory to/from user-space. The overhead from this is hard to tell and >depends on the workload - it can be anything from no visible overhead >to 20-30% overhead. A good rule of thumb is to count with a runtime >overhead of 20%. Whilst you can construct a microbenchmark to show such a large degradation, in real-world applications, the overhead is much lower. That help-text needs updating. You can probably identify some workloads that 4g4g significantly helps. But for most uses on the desktop it is not needed. I am just suggesting that the default kernel doesn't need to be 4g4g. Or at least I am not aware of any reason for it to be so. Perhaps we could agree to a i686 kernel without 4g4g being installable from the CD as an option during installation. theres actually been some internal discussion over the last week or so whether or not to ship it as default any more. For FC4, its likely that the default kernel wont be 4g/4g for a number of reasons. I'm also pondering doing the same to the next FC2/FC3 kernel update. *** Bug 131037 has been marked as a duplicate of this bug. *** *** Bug 124251 has been marked as a duplicate of this bug. *** *** Bug 119838 has been marked as a duplicate of this bug. *** *** Bug 144023 has been marked as a duplicate of this bug. *** *** Bug 145696 has been marked as a duplicate of this bug. *** *** Bug 147250 has been marked as a duplicate of this bug. *** Thanks, Warren, for steering me to this bug report. I'd like to report that the kernel in http://www.sandersweb.net/david/virtualpc/fedora3.zip recommended above did *not* work for me, but the kernel in the i586 RPM did work. Brian, the fedira3.zip is only applicable to Fedora Core 3-Test 3. You should not be using it on the final Core 3 code. See vpc.visualwin.com for instruction related to the final release. *** Bug 158385 has been marked as a duplicate of this bug. *** |