[U-Boot] Build error in davinci_emac driver
Sandeep,
Few of your commits will not allow using davinci_emac driver for non-davinci family of devices. For example,
commit 840f8923a076b02b983cf58c9466b5bdf5957133 Author: Sandeep Paulraj s-paulraj@ti.com Date: Tue Dec 28 15:43:16 2010 -0500
DaVinci DM6467: Added ET1011C (LSI) PHY support
This adds new file, +COBJS-$(CONFIG_DRIVER_TI_EMAC) += lxt972.o dp83848.o et1011c.o
And also modifies the davinci_emac.x file for new API added in et1011c.c.
Ideally all the platform specific drivers should be part of some callback API and should not be hardcoded to driver file.
We will have to clean up this implementation, let me submit all other trivial patches first then I will come to this.
Thanks, Vaibhav
Sandeep,
Few of your commits will not allow using davinci_emac driver for non- davinci family of devices. For example,
This has happened in the kernel also. How is a developer supposed to know which new devices are in the pipeline.
Infact we recently had a case in the kernel NAND driver, where the same IP was being used in a DSP. Even after so much cleanup in the Davinci NAND driver we figured that the driver was specific to devices with an ARM and so I had to submit a patch which would result in it working on devices with only a DSP.
commit 840f8923a076b02b983cf58c9466b5bdf5957133 Author: Sandeep Paulraj s-paulraj@ti.com Date: Tue Dec 28 15:43:16 2010 -0500
DaVinci DM6467: Added ET1011C (LSI) PHY supportThis adds new file, +COBJS-$(CONFIG_DRIVER_TI_EMAC) += lxt972.o dp83848.o et1011c.o
We will have to clean up this implementation, let me submit all other trivial patches first then I will come to this.
Improvements are always welcome. You can submit a patch to mailing list.
--Sandeep
-----Original Message----- From: Paulraj, Sandeep Sent: Tuesday, August 02, 2011 9:19 PM To: Hiremath, Vaibhav Cc: U-Boot@lists.denx.de Subject: RE: Build error in davinci_emac driver
Sandeep,
Few of your commits will not allow using davinci_emac driver for non- davinci family of devices. For example,
This has happened in the kernel also.
[Hiremath, Vaibhav] I am not saying kernel is 100% clean, there could be issues there as well.
How is a developer supposed to know which new devices are in the pipeline.
[Hiremath, Vaibhav] There is NO need to know about this, we can very well implement/design considering other devices.
Infact we recently had a case in the kernel NAND driver, where the same IP was being used in a DSP. Even after so much cleanup in the Davinci NAND driver we figured that the driver was specific to devices with an ARM and so I had to submit a patch which would result in it working on devices with only a DSP.
[Hiremath, Vaibhav] I am not sure about this; if it is the case then definitely it required fix/cleanup.
commit 840f8923a076b02b983cf58c9466b5bdf5957133 Author: Sandeep Paulraj s-paulraj@ti.com Date: Tue Dec 28 15:43:16 2010 -0500
DaVinci DM6467: Added ET1011C (LSI) PHY supportThis adds new file, +COBJS-$(CONFIG_DRIVER_TI_EMAC) += lxt972.o dp83848.o et1011c.o
We will have to clean up this implementation, let me submit all other trivial patches first then I will come to this.
Improvements are always welcome. You can submit a patch to mailing list.
[Hiremath, Vaibhav] Thanks. Will do the same.
Sandeep, Have you reviewed other patches which I have submitted to the list? If not can you review and merge it for your next pull request?
Thanks, Vaibhav
--Sandeep
participants (2)
-
Hiremath, Vaibhav -
Paulraj, Sandeep