On Thu, Sep 06, 2018 at 09:38:26AM +0200, Lukas Wunner wrote: > On Wed, Sep 05, 2018 at 05:30:16PM -0500, Bjorn Helgaas wrote: > > > > > On 8/19/2018 10:29 AM, Lukas Wunner wrote: > > > > > > Simplify the driver by merging struct slot into struct controller. > > > > > > Merge the slot constructor pcie_init_slot() and the destructor > > > > > > pcie_cleanup_slot() into the controller counterparts. > > > > > > Rename the slot struct's "lock" member to "state_lock" and rename > > > > > > "work" to "button_work" for clarity. > > > > Could the rename be done in a separate patch before the unification? > > If so, that would be nice. > > Okay, I've already reworked my internal development branch to do the > unification first, then the rename, but can easily turn the order around, > no problem at all. I don't care which order they're in. I just wasn't sure if the unification was possible without the renaming. > The preceding patches 1 to 3 of this series could be merged independently, > or would you prefer me to resend them? The first one is also included in > Keith's error handling rework series (posted yesterday): > > [1/9] PCI: Simplify disconnected marking > https://patchwork.ozlabs.org/patch/959373/ > [2/9] PCI: pciehp: Drop unnecessary includes > https://patchwork.ozlabs.org/patch/959376/ > [3/9] [3/9] PCI: pciehp: Drop hotplug_slot_ops wrappers > https://patchwork.ozlabs.org/patch/959377/ I applied patches 1-3 to pci/hotplug. If you could rebase what's left on top of that, it would be great. > Important side note: Keith's series includes a pciehp fix which should > probably go into 4.19: > > [PATCHv2,15/20] PCI/pciehp: Fix powerfault detection order > https://patchwork.ozlabs.org/patch/966678/ Thanks for the heads-up; I probably wouldn't have picked that out. > Now, if that patch goes into, say, 4.19-rc3, the pci/hotplug branch would > have to be rebased on top of 4.19-rc3 to avoid conflicts during the 4.20 > merge window. I'll try to remember to do this. I have no problem rebasing my branches to -rc3. Bjorn
