On Tue, 2008-06-17 at 20:01 +0000, Gary Hawco wrote: > I was reviewing your patches from 9 hours ago and was confused as to > whether delalloc still needs writeback mode. I see the > enable-delalloc-by-default-patch that quickly followed the > delalloc-ext4-writeback mode. Does this mean that data=writeback mode must > be used, and if this condition is met, then delalloc is enabled, or is > delalloc enabled even if ordered data mode? Any clarification on this would > be most appreciated. > Hi Gary, thanks for looking into this. Currently, if you apply the whole the patch queue patches, you should have a ext4 fs with delalloc over ordered mode by default. Previously without the delayed allocation support for ordered mode, the patch queue enabled delalloc over writeback mode by default. To clarify, the delayed allocation patches in the series are: # New delayed allocation patch delalloc-vfs.patch delalloc-ext4-writeback-mode.patch #delalloc support for reverse locking Add-range_cont-mode-for-writeback.patch delalloc-ext4-reserve-locking-order-support.patch #delalloc block reservation(ENOSPC) handling percpucounter-add-sum-and-set-function.patch delalloc-ext4-ENOSPC-handling.patch #delalloc new ordered mode delalloc-new-ordered-mode.patch enable-delalloc-by-default.patch show-delalloc-option.patch So, with delalloc-ext4-writeback-mode.patch, it has delalloc support for writeback mode, but you need to enable this feature clearly with mount option -o delalloc, data=writeback, With delalloc-new-ordered-mode.patch, it add delalloc support for ordered mode (the new ordered mode, patches in the patch queue). If you only apply to here, you need to enable delalloc on ordered mode by mount -o delalloc, data=ordered With the enable-delalloc-by-default.patch, it enables delalloc over new ordered mode by default (no need to specifying delalloc and ordered mount options), which exercise the new ordered code and also delayed allocation both. Mingming -- To unsubscribe from this list: send the line "unsubscribe linux-ext4" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
