Bug#463606: confirmed: rebuild 2.6.26-1-686 as CONFIG_X86_GENERIC resolves

2008-08-29 Thread Simon Mackinlay
${Subject} - rebuilding debian kernel per... debian/config/i386/config.686: CONFIG_X86_GENERIC=y rm debian/abi/2.6.26-1/i386_none_686 <-- obviously not ideal, a quick hack to prove the point and installing the result boots just fine (again still using the noreplace-paravirt boot commandline).

Bug#463606: still present in 2.6.26-1-686

2008-08-29 Thread Simon Mackinlay
${Subject} - also - wrt: getting the stock -486- image to boot, add "noreplace-paravirt" to the boot command line. Recompiling -686- with CONFIG_X86_GENERIC=y ought to address the Int 6 at early boot; is this a possibility? -- Nothing says Labor Day like 500hp of American muscle Visit OnCars.com

Bug#355881: initramfs-tools: should be hook-functions helper for firmware?

2007-01-16 Thread Simon Mackinlay
> sure but they look very differently from what you think. > the support needs to land in modprobe first. Yep fair call, just noticed the discrepancy between the way we handle modules ... > nothing to be seen there, just use copy_exec ... against that for firmware; I haven't got religion one way

Bug#406763: firmware-ipw3945: no initramfs-tools hook script

2007-01-16 Thread Simon Mackinlay
> why? Just saw the resulting error and reporting it and cause seemed the logical thing to do... > nobody wants rootnfs over wlan. ... if this is the only reason we attempt to bring up networking this early in boot I guess it's a non-issue as you suggest; I originally misdiagnosed this locally a

Bug#406769: linux-image-2.6.18-3-amd64: rdtsc instruction does not serialise on core2 family

2007-01-13 Thread Simon Mackinlay
Package: linux-image-2.6.18-3-amd64 Version: 2.6.18-8 Severity: critical Justification: breaks the whole system Commit from 2.6.18.6 upstream is vitally important for Core 2 family. Without this TSC drift between cores will result in unpredictable hard locks and reboots. commit e4a835d383dc582

Bug#406767: linux-image-2.6-3-amd64: rdtsc instruction does not serialise on core2 cpu's

2007-01-13 Thread Simon Mackinlay
Package: linux-image-2.6-3-amd64 Severity: critical Justification: breaks the whole system Commit from 2.6.18.6 upstream is vitally important for Core 2 family. Without this TSC drift between cores will result in unpredictable hard locks and reboots. commit e4a835d383dc58212a9648ef905cb8087e0c4

Bug#355881: initramfs-tools: should be hook-functions helper for firmware?

2007-01-13 Thread Simon Mackinlay
Package: initramfs-tools Version: 0.85e Followup-For: Bug #355881 Perhaps more generically there should be firmware support in /usr/share/initramfs-tools/hook-functions orthogonal to the support for modules. firmware-nonfree would benefit from this in particular; see bts #406763, #386172. -- Pa

Bug#406763: firmware-ipw3945: possible initramfs-tools wishlist item

2007-01-13 Thread Simon Mackinlay
Package: firmware-ipw3945 Version: 0.3 Followup-For: Bug #406763 The thought occurs that firmware-nonfree could benefit from a helper function specifically for firwmare in hook-functions; have hijacked bts #406763 with some comments to this end. -- System Information: Debian Release: 4.0 APT p

Bug#406763: firmware-ipw3945: no initramfs-tools hook script

2007-01-13 Thread Simon Mackinlay
Package: firmware-ipw3945 Version: 0.3 Severity: wishlist Package ought to contain an initramfs-tools hook script so that the firmware image ends up installed into initrd - for example a minimal /usr/share/initramfs-tools/hooks/firmware-ipw3945 is (hopefully) attached. -- System Information: Deb