Rename backend modules to be less generic.
netbk, blkbk and pciback are all rather generic, rename then along the lines of xen-*back.
Split up into several patches since each backend seem to have its own branch.
PATCH 1/3 xen: rename blkbk module xen-blkback.
From: Ian Campbell <[email protected]>
blkbk is rather generic for a modular distro style kernel.
--- drivers/xen/blkback/Makefile | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-)
drivers/xen/blkback/Makefile +2 -2
@@ -1,3 +1,3 @@ -obj-$(CONFIG_XEN_BLKDEV_BACKEND) := blkbk.o +obj-$(CONFIG_XEN_BLKDEV_BACKEND) := xen-blkback.o -blkbk-y := blkback.o xenbus.o interface.o vbd.o +xen-blkback-y := blkback.o xenbus.o interface.o vbd.o
Signature · 1 line
1.6.5.3
PATCH 2/3 xen: rename netbk module xen-netback.
In reply to [email protected]From: Ian Campbell <[email protected]>
netbk is rather generic for a modular distro style kernel.
--- drivers/xen/netback/Makefile | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-)
drivers/xen/netback/Makefile +2 -2
@@ -1,3 +1,3 @@ -obj-$(CONFIG_XEN_NETDEV_BACKEND) := netbk.o +obj-$(CONFIG_XEN_NETDEV_BACKEND) := xen-netback.o -netbk-y := netback.o xenbus.o interface.o +xen-netback-y := netback.o xenbus.o interface.o
Signature · 1 line
1.6.5.3
PATCH 3/3 xen: rename pciback module to xen-pciback.
In reply to [email protected]From: Ian Campbell <[email protected]>
pciback is rather generic for a modular distro style kernel.
--- drivers/xen/pciback/Makefile | 24 ++++++++++++------------ 1 files changed, 12 insertions(+), 12 deletions(-)
drivers/xen/pciback/Makefile +12 -12
@@ -1,16 +1,16 @@ -obj-$(CONFIG_XEN_PCIDEV_BACKEND) += pciback.o +obj-$(CONFIG_XEN_PCIDEV_BACKEND) += xen-pciback.o -pciback-y := pci_stub.o pciback_ops.o xenbus.o -pciback-y += conf_space.o conf_space_header.o \ - conf_space_capability.o \ - conf_space_capability_vpd.o \ - conf_space_capability_pm.o \ - conf_space_quirks.o -pciback-$(CONFIG_PCI_MSI) += conf_space_capability_msi.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_VPCI) += vpci.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_SLOT) += slot.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_PASS) += passthrough.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER) += controller.o +xen-pciback-y := pci_stub.o pciback_ops.o xenbus.o +xen-pciback-y += conf_space.o conf_space_header.o \ + conf_space_capability.o \ + conf_space_capability_vpd.o \ + conf_space_capability_pm.o \ + conf_space_quirks.o +xen-pciback-$(CONFIG_PCI_MSI) += conf_space_capability_msi.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_VPCI) += vpci.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_SLOT) += slot.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_PASS) += passthrough.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER) += controller.o ifeq ($(CONFIG_XEN_PCIDEV_BE_DEBUG),y) EXTRA_CFLAGS += -DDEBUG
Signature · 1 line
1.6.5.3
Re: [PATCH 3/3] xen: rename pciback module to xen-pciback.
Quoted 1 line · On Thu, Dec 03, 2009 at 09:56:20PM +0000, [email protected] wrote:
pciback is rather generic for a modular distro style kernel.
Do you have a patch for Xen python stack to check ownership of PCI devices to also be 'xen-pciback'?
Quoted 4 lines
Signed-off-by Ian Campbell <[email protected]>Cc Jeremy Fitzhardinge <[email protected]>Cc Konrad Rzeszutek Wilk <[email protected]>
Quoted 40 lines
--- drivers/xen/pciback/Makefile | 24 ++++++++++++------------ 1 files changed, 12 insertions(+), 12 deletions(-)drivers/xen/pciback/Makefile +12
-12@@ -1,16 +1,16 @@ -obj-$(CONFIG_XEN_PCIDEV_BACKEND) += pciback.o +obj-$(CONFIG_XEN_PCIDEV_BACKEND) += xen-pciback.o -pciback-y := pci_stub.o pciback_ops.o xenbus.o -pciback-y += conf_space.o conf_space_header.o \ - conf_space_capability.o \ - conf_space_capability_vpd.o \ - conf_space_capability_pm.o \ - conf_space_quirks.o -pciback-$(CONFIG_PCI_MSI) += conf_space_capability_msi.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_VPCI) += vpci.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_SLOT) += slot.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_PASS) += passthrough.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER) += controller.o +xen-pciback-y := pci_stub.o pciback_ops.o xenbus.o +xen-pciback-y += conf_space.o conf_space_header.o \ + conf_space_capability.o \ + conf_space_capability_vpd.o \ + conf_space_capability_pm.o \ + conf_space_quirks.o +xen-pciback-$(CONFIG_PCI_MSI) += conf_space_capability_msi.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_VPCI) += vpci.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_SLOT) += slot.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_PASS) += passthrough.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER) += controller.o ifeq ($(CONFIG_XEN_PCIDEV_BE_DEBUG),y) EXTRA_CFLAGS += -DDEBUGSignature · 1 line
1.6.5.3
Re: [PATCH 3/3] xen: rename pciback module to xen-pciback.
Quoted 7 lines · Konrad Rzeszutek Wilk
Quoted 2 lines · On Thu, Dec 03, 2009 at 09:56:20PM +0000, [email protected] wrote:
pciback is rather generic for a modular distro style kernel.
Do you have a patch for Xen python stack to check ownership of PCI devices to also be 'xen-pciback'?
Is the module name the same as the driver name?
J
Re: [PATCH 3/3] xen: rename pciback module to xen-pciback.
Quoted 10 lines · Jeremy Fitzhardinge
Quoted 7 lines · Konrad Rzeszutek Wilk
Quoted 2 lines · On Thu, Dec 03, 2009 at 09:56:20PM +0000, [email protected] wrote:
pciback is rather generic for a modular distro style kernel.
Do you have a patch for Xen python stack to check ownership of PCI devices to also be 'xen-pciback'?
Is the module name the same as the driver name?
Not with this patch. The driver name is 'pciback' but the module name will be 'xen-pciback'. I am not sure how that is going to resolve in SysFS - hopefully it will remain 'pciback' for the pertient directories.
Re: [PATCH 3/3] xen: rename pciback module to xen-pciback.
In reply to Konrad Rzeszutek WilkQuoted 5 lines · Konrad Rzeszutek Wilk
Quoted 1 line · On Thu, Dec 03, 2009 at 09:56:20PM +0000, [email protected] wrote:
pciback is rather generic for a modular distro style kernel.
Do you have a patch for Xen python stack to check ownership of PCI devices to also be 'xen-pciback'?
Ah, it never even occurred to me that such a thing would be necessary.
I think (but I should check) that the sysfs name will remain pciback since it comes from the .name in struct xenbus_driver. I'm not sure if it is confusing or not to have the module name be different to the driver name nor how common it is.
Ian.
Quoted 48 lines
Quoted 5 lines
Signed-off-by Ian Campbell <[email protected]>Cc Jeremy Fitzhardinge <[email protected]>Cc Konrad Rzeszutek Wilk <[email protected]>Quoted 40 lines
--- drivers/xen/pciback/Makefile | 24 ++++++++++++------------ 1 files changed, 12 insertions(+), 12 deletions(-)drivers/xen/pciback/Makefile +12
-12@@ -1,16 +1,16 @@ -obj-$(CONFIG_XEN_PCIDEV_BACKEND) += pciback.o +obj-$(CONFIG_XEN_PCIDEV_BACKEND) += xen-pciback.o -pciback-y := pci_stub.o pciback_ops.o xenbus.o -pciback-y += conf_space.o conf_space_header.o \ - conf_space_capability.o \ - conf_space_capability_vpd.o \ - conf_space_capability_pm.o \ - conf_space_quirks.o -pciback-$(CONFIG_PCI_MSI) += conf_space_capability_msi.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_VPCI) += vpci.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_SLOT) += slot.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_PASS) += passthrough.o -pciback-$(CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER) += controller.o +xen-pciback-y := pci_stub.o pciback_ops.o xenbus.o +xen-pciback-y += conf_space.o conf_space_header.o \ + conf_space_capability.o \ + conf_space_capability_vpd.o \ + conf_space_capability_pm.o \ + conf_space_quirks.o +xen-pciback-$(CONFIG_PCI_MSI) += conf_space_capability_msi.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_VPCI) += vpci.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_SLOT) += slot.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_PASS) += passthrough.o +xen-pciback-$(CONFIG_XEN_PCIDEV_BACKEND_CONTROLLER) += controller.o ifeq ($(CONFIG_XEN_PCIDEV_BE_DEBUG),y) EXTRA_CFLAGS += -DDEBUGSignature · 1 line
1.6.5.3
Re: [PATCH 3/3] xen: rename pciback module to xen-pciback.
Quoted 7 lines · Ian Campbell
Ah, it never even occurred to me that such a thing would be necessary.
I think (but I should check) that the sysfs name will remain pciback since it comes from the .name in struct
xenbus_driver. I'm not sure if it is confusing or not to have the module name be different to the driver name nor how common it is.
The driver name is a bit of an ABI for hotplug, etc, anyway, so having the Xen tools depend on that name isn't very odd. And while it is best if the module matches the driver name, it's not uncommon for it to mismatch - and we've been doing it for ages anyway (xen_netfront is called "vif" and aliased "xen:vif" and "xennet").
J