Re: [PATCH] dmaengine: Consistently define pci_device_ids using named initializers

[Uwe Kleine-König (The Capable Hub)]

Hello Andy,

On Mon, May 04, 2026 at 05:09:55PM +0300, Andy Shevchenko wrote:
> On Mon, May 04, 2026 at 03:55:00PM +0200, Uwe Kleine-König (The Capable Hub) wrote:
> > On Mon, May 04, 2026 at 01:29:12PM +0300, Andy Shevchenko wrote:
> > > On Mon, May 04, 2026 at 12:20:06PM +0200, Uwe Kleine-König (The Capable Hub) wrote:
> > > > The .driver_data member of the various struct pci_device_id arrays were
> > > > initialized by list expressions. This isn't easily readable if you're
> > > > not into PCI. Using named initializers is more explicit and thus easier
> > > > to parse. Also skip explicit assignments of 0 (which the compiler then
> > > > takes care of).
> > > > 
> > > > This change doesn't introduce changes to the compiled pci_device_id
> > > > arrays. Tested on x86 and arm64.
> > > 
> > > HSU driver has different change ("Also" is a strong sign to the split required).
> > 
> > HSU is in the category "skip explicit assignments of 0", so I think
> > that's fine. I could be talked into splitting if that's what is wanted.
> 
> Yes, please. I will Rb/Ack it immediately when standalone change.
> 
> ...
> 
> > > >  static const struct pci_device_id pch_dma_id_table[] = {
> > > > -	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_EG20T_PCH_DMA_8CH), 8 },
> > > > -	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_EG20T_PCH_DMA_4CH), 4 },
> > > > -	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7213_DMA1_8CH), 8}, /* UART Video */
> > > > -	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7213_DMA2_8CH), 8}, /* PCMIF SPI */
> > > > -	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7213_DMA3_4CH), 4}, /* FPGA */
> > > > -	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7213_DMA4_12CH), 12}, /* I2S */
> > > > -	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA1_4CH), 4}, /* UART */
> > > > -	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA2_4CH), 4}, /* Video SPI */
> > > > -	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA3_4CH), 4}, /* Security */
> > > > -	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA4_4CH), 4}, /* FPGA */
> > > > -	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7831_DMA1_8CH), 8}, /* UART */
> > > > -	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7831_DMA2_4CH), 4}, /* SPI */
> > > > -	{ 0, },
> > > > +	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_EG20T_PCH_DMA_8CH), .driver_data = 8 },
> > > > +	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_EG20T_PCH_DMA_4CH), .driver_data = 4 },
> > > > +	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7213_DMA1_8CH), .driver_data = 8 },		/* UART Video */
> > > > +	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7213_DMA2_8CH), .driver_data = 8 },		/* PCMIF SPI */
> > > > +	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7213_DMA3_4CH), .driver_data = 4 },		/* FPGA */
> > > > +	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7213_DMA4_12CH), .driver_data = 12 },	/* I2S */
> > > > +	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA1_4CH), .driver_data = 4 },		/* UART */
> > > > +	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA2_4CH), .driver_data = 4 },		/* Video SPI */
> > > > +	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA3_4CH), .driver_data = 4 },		/* Security */
> > > > +	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7223_DMA4_4CH), .driver_data = 4 },		/* FPGA */
> > > > +	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7831_DMA1_8CH), .driver_data = 8 },		/* UART */
> > > > +	{ PCI_VDEVICE(ROHM, PCI_DEVICE_ID_ML7831_DMA2_4CH), .driver_data = 4 },		/* SPI */
> > > > +	{ },
> > > >  };
> > > 
> > > Use PCI_DEVICE_DATA() instead. Same may apply to DesignWare, but one needs to
> > > define the device IDs. I think I may help with that.
> > 
> > I'm not a fan of PCI_DEVICE_DATA. While it could indeed be used to
> > shorten the assignments here, it's less readable in my opinion.
> 
> I'm not fun of these long unreadable lines with tons of repetitions :-)

Seems to be subjective.

> > Compare
> > 
> > 	{ PCI_VDEVICE(INTEL, PCI_DEVICE_ID_EG20T_PCH_DMA_4CH), .driver_data = 4 },
> > 
> > with
> > 
> > 	{ PCI_DEVICE_DATA(INTEL, PCI_DEVICE_ID_EG20T_PCH_DMA_4CH, 4) },
> 
> First of all, with
> 
> 	{ PCI_DEVICE_DATA(INTEL, EG20T_PCH_DMA_4CH, 4) },

Agreed. That doesn't considerably weaken my reasoning however.

> > . For someone who doesn't know what PCI_DEVICE_DATA does, the latter is
> > less understandable.
> 
> Same applicable to many other macros. I don't consider this argument viable.

Also agreed. But other bad macros don't justify using that (admittedly
subjectively) bad PCI_DEVICE_DATA macro that mixes device identity
(.vendor, .device, .subvendor and .subdevice) with a driver specific
struct member.

> > Also PCI_DEVICE_DATA has a cast which is something I want to get rid of.
> 
> Yes, and you will get rid of in one place instead of tons of them.

This would require another (subjectively bad) macro PCI_DEVICE_DATAPTR.
I think I let someone else tackle that quest.

Best regards
Uwe

Attachment: signature.asc
Description: PGP signature