[v3,16/37] drivers: remove references to make config options
Checks
Commit Message
Make is no longer supported for compiling DPDK, references are now
removed in code comments.
Signed-off-by: Ciara Power <ciara.power@intel.com>
---
drivers/bus/pci/linux/pci_vfio.c | 1 -
drivers/net/i40e/i40e_rxtx.c | 1 -
2 files changed, 2 deletions(-)
Comments
On Thu Sep 3, 2020 at 6:26 PM CEST, Ciara Power wrote:
> Make is no longer supported for compiling DPDK, references are now
> removed in code comments.
It should be squashed with the patch removing the config options.
In general, I feel this patchset would benefit a bit more squashing.
On Sun, Sep 06, 2020 at 12:24:49PM +0200, Thomas Monjalon wrote:
> On Thu Sep 3, 2020 at 6:26 PM CEST, Ciara Power wrote:
> > Make is no longer supported for compiling DPDK, references are now
> > removed in code comments.
>
> It should be squashed with the patch removing the config options.
>
> In general, I feel this patchset would benefit a bit more squashing.
Yes, but that makes it harder to review and work with. For example, for
removing the config options each document needs to be modified to remove
references to those, and once make itself is removed again all docs need to
be modified. To review properly, the doc needs to be scanned to check no
references, especially indirect references, are missed. Given that e.g. the
NIC guide doc is 55 chapters, that's a whole lot of reviewing you want to
minimize, so having all changes to that one doc in one patch is much more
feasible.
Really, removing the whole build system could all be done in one patch, as
removing parts a bit at a time doesn't really make sense as it's all broken
once one part is gone. However, the resulting patch would be enormous, so I
suggest keeping the parts separate for review and then squash on merge if
so desired.
/Bruce
07/09/2020 10:18, Bruce Richardson:
> On Sun, Sep 06, 2020 at 12:24:49PM +0200, Thomas Monjalon wrote:
> > On Thu Sep 3, 2020 at 6:26 PM CEST, Ciara Power wrote:
> > > Make is no longer supported for compiling DPDK, references are now
> > > removed in code comments.
> >
> > It should be squashed with the patch removing the config options.
> >
> > In general, I feel this patchset would benefit a bit more squashing.
>
> Yes, but that makes it harder to review and work with. For example, for
> removing the config options each document needs to be modified to remove
> references to those, and once make itself is removed again all docs need to
> be modified. To review properly, the doc needs to be scanned to check no
> references, especially indirect references, are missed. Given that e.g. the
> NIC guide doc is 55 chapters, that's a whole lot of reviewing you want to
> minimize, so having all changes to that one doc in one patch is much more
> feasible.
>
> Really, removing the whole build system could all be done in one patch, as
> removing parts a bit at a time doesn't really make sense as it's all broken
> once one part is gone. However, the resulting patch would be enormous, so I
> suggest keeping the parts separate for review and then squash on merge if
> so desired.
I will keep doc patches separate because it's too big and requires more reviews.
The separation between makefiles and configs is also fine.
@@ -34,7 +34,6 @@
* This code tries to determine if the PCI device is bound to VFIO driver,
* and initialize it (map BARs, set up interrupts) if that's the case.
*
- * This file is only compiled if CONFIG_RTE_EAL_VFIO is set to "y".
*/
#ifdef VFIO_PRESENT
@@ -3438,7 +3438,6 @@ i40e_set_default_pctype_table(struct rte_eth_dev *dev)
}
#ifndef RTE_LIBRTE_I40E_INC_VECTOR
-/* Stubs needed for linkage when CONFIG_RTE_LIBRTE_I40E_INC_VECTOR is set to 'n' */
int
i40e_rx_vec_dev_conf_condition_check(struct rte_eth_dev __rte_unused *dev)
{