Control: tag -1 - moreinfo Control: tag -1 pending I got fed up of seeing ia64 build failures so I've re-added ia64 configurations on the master branch. On Wed, 2018-01-17 at 01:20 +0000, Ben Hutchings wrote: > Control: tag -1 moreinfo > > On Mon, 2018-01-08 at 19:04 -0500, Jason Duerstock wrote: > > Source: linux > > Severity: normal > > Tags: patch > > > > Dear Maintainer, > > > > As you may be aware, the ia64 architecture has recently been added > > back to Debian, but now resides in Debian ports. > > The attached patch should enable the linux package to build the ia64 > > kernel again. > > > > Thanks for your time! > > This appears to be almost exactly reverting the change I made to remove > ia64 support, which is not the right thing to do. > > You need to actually review the changes that have happened in the 2.5 > years since then and update the config accordingly. In particular, the > following symbols no longer exist in Linux 4.15-rc8: > > CONFIG_BLK_CPQ_CISS_DA > CONFIG_BLK_CPQ_DA > CONFIG_CISS_SCSI_TAPE > CONFIG_I2O > CONFIG_I2O_BLOCK > CONFIG_I2O_CONFIG > CONFIG_I2O_PROC > CONFIG_I2O_SCSI > CONFIG_MMTIMER > CONFIG_SCSI_DC390T I deleted all of these from the configuration. > The commit message for the removal of CONFIG_MMTIMER upstream (commit > 07903ada96139ced48f2f893fe57a26a8fbc6043) implies that SGI SN2 systems > are no longer supported, in which case presumably the sn-modules udeb > should also be removed. I was confused about this; there evidently is still general support for SN2 systems. > Do Itanium systems typically have floppy drives? If not, delete the > "suggests: fdutils" from debian/config/ia64/defines. I've done this. > Shouldn't the "mckinley" configuration be renamed, since it's supposed > to support later processors as well? > > Does it still make sense to build an "itanium" configuration, given how > few Merced systems exist? I've left the flavour names as they are but changed their descriptions. > Also, do you have any idea whether these bugs have been fixed upstream: > > https://bugs.debian.org/679545 This seems to be unfixed (the patches weren't applied upstream). > https://bugs.debian.org/691576 This seems to have been a gcc bug that is now fixed. > https://bugs.debian.org/728706 Unknown. > If not, those should be reopened when ia64 is enabled again. Ben. -- Ben Hutchings Quantity is no substitute for quality, but it's the only one we've got.
Attachment:
signature.asc
Description: This is a digitally signed message part