Linux USB
 help / color / mirror / Atom feed
From: Sakari Ailus <sakari.ailus@linux.intel.com>
To: Niklas Neronin <niklas.neronin@linux.intel.com>
Cc: mathias.nyman@linux.intel.com, linux-usb@vger.kernel.org, "Rai,
	Amardeep" <amardeep.rai@intel.com>
Subject: Re: [PATCH 11/13] usb: xhci: standardize single bit-field macros
Date: Tue, 11 Nov 2025 10:51:41 +0200	[thread overview]
Message-ID: <aRL5HQw2-a94FPpA@kekkonen.localdomain> (raw)
In-Reply-To: <20251110151450.635410-12-niklas.neronin@linux.intel.com>

Hi Niklas,

Thanks for this!

On Mon, Nov 10, 2025 at 04:14:48PM +0100, Niklas Neronin wrote:
> Convert single bit-field macros to simple masks. The change makes the
> masks more universal. Multi bit-field macros are changed in the next
> commit. After both changes, all masks in xhci-caps.h will follow the
> same format. I plan to introduce this change to all xhci macros.
> 
> Bit shift operations on a 32-bit signed can be problematic on some
> architectures. Instead use BIT() macro, which returns a 64-bit unsigned
> value. This ensures that the shift operation is performed on an unsigned
> type, which is safer and more portable across different architectures.
> Using unsigned integers for bit shifts avoids issues related to sign bits
> and ensures consistent behavior.
> 
> Switch from 32-bit to 64-bit?
> As far as I am aware, this does not cause any issues.
> Performing bitwise operations between 32 and 64 bit values, the smaller
> operand is promoted to match the size of the larger one, resulting in a
> 64-bit operation. This promotion extends the 32-bit value to 64 bits,
> by zero-padding (for unsigned).
> 
> Will the change to 64-bit slow down the xhci driver?
> On a 64-bit architecture - No. On a 32-bit architecture, yes? but in my
> opinion the performance decrease does not outweigh the readability and
> other benefits of using BIT() macro.
> 
> Why not use FIELD_GET() and FIELD_PREP()?
> While they can be used for single bit macros, I prefer to use simple
> bitwise operation directly. Because, it takes less space, is less overhead
> and is as clear as if using FIELD_GET() and FIELD_PREP().
> 
> Why not use test_bit() macro?
> Same reason as with FIELD_GET() and FIELD_PREP().
> 
> Suggested-by: Sakari Ailus <sakari.ailus@linux.intel.com>
> Signed-off-by: Niklas Neronin <niklas.neronin@linux.intel.com>

Reviewed-by: Sakari Ailus <sakari.ailus@linux.intel.com>

> ---
>  drivers/usb/host/xhci-caps.h    | 50 +++++++++++++++++----------------
>  drivers/usb/host/xhci-debugfs.c |  2 +-
>  drivers/usb/host/xhci-hub.c     |  6 ++--
>  drivers/usb/host/xhci-mem.c     |  7 ++---
>  drivers/usb/host/xhci-ring.c    |  8 +++---
>  drivers/usb/host/xhci-trace.h   |  2 +-
>  drivers/usb/host/xhci.c         |  2 +-
>  7 files changed, 39 insertions(+), 38 deletions(-)
> 
> diff --git a/drivers/usb/host/xhci-caps.h b/drivers/usb/host/xhci-caps.h
> index 99557df89f88..89dd46767885 100644
> --- a/drivers/usb/host/xhci-caps.h
> +++ b/drivers/usb/host/xhci-caps.h
> @@ -4,6 +4,8 @@
>   * xHCI Specification Section 5.3, Revision 1.2.
>   */
>  
> +#include <linux/bits.h>
> +
>  /* hc_capbase - bitmasks */
>  /* bits 7:0 - Capability Registers Length */
>  #define HC_LENGTH(p)		((p) & 0xff)
> @@ -32,7 +34,7 @@
>   * xHCI specification section 5.3.4.
>   */
>  #define HCS_IST_VALUE(p)	((p) & 0x7)
> -#define HCS_IST_UNIT(p)		((p) & (1 << 3))
> +#define HCS_IST_UNIT		BIT(3)
>  /* bits 7:4 - Event Ring Segment Table Max, 2^(n) */
>  #define HCS_ERST_MAX(p)		(((p) >> 4) & 0xf)
>  /* bits 20:8 - Rsvd */
> @@ -52,30 +54,30 @@
>  
>  /* HCCPARAMS1 - hcc_params - bitmasks */
>  /* bit 0 - 64-bit Addressing Capability */
> -#define HCC_64BIT_ADDR(p)	((p) & (1 << 0))
> +#define HCC_64BIT_ADDR		BIT(0)
>  /* bit 1 - BW Negotiation Capability */
> -#define HCC_BANDWIDTH_NEG(p)	((p) & (1 << 1))
> +#define HCC_BANDWIDTH_NEG	BIT(1)
>  /* bit 2 - Context Size */
> -#define HCC_64BYTE_CONTEXT(p)	((p) & (1 << 2))
> -#define CTX_SIZE(_hcc)		(HCC_64BYTE_CONTEXT(_hcc) ? 64 : 32)
> +#define HCC_64BYTE_CONTEXT	BIT(2)
> +#define CTX_SIZE(_hcc)		(_hcc & HCC_64BYTE_CONTEXT ? 64 : 32)
>  /* bit 3 - Port Power Control */
> -#define HCC_PPC(p)		((p) & (1 << 3))
> +#define HCC_PPC			BIT(3)
>  /* bit 4 - Port Indicators */
> -#define HCS_INDICATOR(p)	((p) & (1 << 4))
> +#define HCS_INDICATOR		BIT(4)
>  /* bit 5 - Light HC Reset Capability */
> -#define HCC_LIGHT_RESET(p)	((p) & (1 << 5))
> +#define HCC_LIGHT_RESET		BIT(5)
>  /* bit 6 - Latency Tolerance Messaging Capability */
> -#define HCC_LTC(p)		((p) & (1 << 6))
> +#define HCC_LTC			BIT(6)
>  /* bit 7 - No Secondary Stream ID Support */
> -#define HCC_NSS(p)		((p) & (1 << 7))
> +#define HCC_NSS			BIT(7)
>  /* bit 8 - Parse All Event Data */
>  /* bit 9 - Short Packet Capability */
> -#define HCC_SPC(p)		((p) & (1 << 9))
> +#define HCC_SPC			BIT(9)
>  /* bit 10 - Stopped EDTLA Capability */
>  /* bit 11 - Contiguous Frame ID Capability */
> -#define HCC_CFC(p)		((p) & (1 << 11))
> +#define HCC_CFC			BIT(11)
>  /* bits 15:12 - Max size for Primary Stream Arrays, 2^(n+1) */
> -#define HCC_MAX_PSA(p)		(1 << ((((p) >> 12) & 0xf) + 1))
> +#define HCC_MAX_PSA(p)		BIT(((((p) >> 12) & 0xf) + 1))
>  /* bits 31:16 - xHCI Extended Capabilities Pointer, from PCI base: 2^(n) */
>  #define HCC_EXT_CAPS(p)		(((p) >> 16) & 0xffff)
>  
> @@ -91,26 +93,26 @@
>  
>  /* HCCPARAMS2 - hcc_params2 - bitmasks */
>  /* bit 0 - U3 Entry Capability */
> -#define	HCC2_U3C(p)		((p) & (1 << 0))
> +#define	HCC2_U3C		BIT(0)
>  /* bit 1 - Configure Endpoint Command Max Exit Latency Too Large Capability */
> -#define	HCC2_CMC(p)		((p) & (1 << 1))
> +#define	HCC2_CMC		BIT(1)
>  /* bit 2 - Force Save Context Capabilitu */
> -#define	HCC2_FSC(p)		((p) & (1 << 2))
> +#define	HCC2_FSC		BIT(2)
>  /* bit 3 - Compliance Transition Capability, false: compliance is enabled by default */
> -#define	HCC2_CTC(p)		((p) & (1 << 3))
> +#define	HCC2_CTC		BIT(3)
>  /* bit 4 - Large ESIT Payload Capability, true: HC support ESIT payload > 48k */
> -#define	HCC2_LEC(p)		((p) & (1 << 4))
> +#define	HCC2_LEC		BIT(4)
>  /* bit 5 - Configuration Information Capability */
> -#define	HCC2_CIC(p)		((p) & (1 << 5))
> +#define	HCC2_CIC		BIT(5)
>  /* bit 6 - Extended TBC Capability, true: Isoc burst count > 65535 */
> -#define	HCC2_ETC(p)		((p) & (1 << 6))
> +#define	HCC2_ETC		BIT(6)
>  /* bit 7 - Extended TBC TRB Status Capability */
> -#define HCC2_ETC_TSC(p)         ((p) & (1 << 7))
> +#define HCC2_ETC_TSC		BIT(7)
>  /* bit 8 - Get/Set Extended Property Capability */
> -#define HCC2_GSC(p)             ((p) & (1 << 8))
> +#define HCC2_GSC		BIT(8)
>  /* bit 9 - Virtualization Based Trusted I/O Capability */
> -#define HCC2_VTC(p)             ((p) & (1 << 9))
> +#define HCC2_VTC		BIT(9)
>  /* bit 10 - Rsvd */
>  /* bit 11 - HC support Double BW on a eUSB2 HS ISOC EP */
> -#define HCC2_EUSB2_DIC(p)       ((p) & (1 << 11))
> +#define HCC2_EUSB2_DIC		BIT(11)
>  /* bits 31:12 - Rsvd */
> diff --git a/drivers/usb/host/xhci-debugfs.c b/drivers/usb/host/xhci-debugfs.c
> index f0475cf8eef8..e45545fa3c66 100644
> --- a/drivers/usb/host/xhci-debugfs.c
> +++ b/drivers/usb/host/xhci-debugfs.c
> @@ -355,7 +355,7 @@ static ssize_t xhci_port_write(struct file *file,  const char __user *ubuf,
>  
>  	if (!strncmp(buf, "compliance", 10)) {
>  		/* If CTC is clear, compliance is enabled by default */
> -		if (!HCC2_CTC(xhci->hcc_params2))
> +		if (!(xhci->hcc_params2 & HCC2_CTC))
>  			return count;
>  		spin_lock_irqsave(&xhci->lock, flags);
>  		/* compliance mode can only be enabled on ports in RxDetect */
> diff --git a/drivers/usb/host/xhci-hub.c b/drivers/usb/host/xhci-hub.c
> index e7c1952cffc5..422028ebce49 100644
> --- a/drivers/usb/host/xhci-hub.c
> +++ b/drivers/usb/host/xhci-hub.c
> @@ -110,7 +110,7 @@ static int xhci_create_usb3x_bos_desc(struct xhci_hcd *xhci, char *buf,
>  	ss_cap->bU2DevExitLat = 0; /* set later */
>  
>  	reg = readl(&xhci->cap_regs->hcc_params);
> -	if (HCC_LTC(reg))
> +	if (reg & HCC_LTC)
>  		ss_cap->bmAttributes |= USB_LTM_SUPPORT;
>  
>  	if ((xhci->quirks & XHCI_LPM_SUPPORT)) {
> @@ -263,7 +263,7 @@ static void xhci_common_hub_descriptor(struct xhci_hcd *xhci,
>  	desc->bNbrPorts = ports;
>  	temp = 0;
>  	/* Bits 1:0 - support per-port power switching, or power always on */
> -	if (HCC_PPC(xhci->hcc_params))
> +	if (xhci->hcc_params & HCC_PPC)
>  		temp |= HUB_CHAR_INDV_PORT_LPSM;
>  	else
>  		temp |= HUB_CHAR_NO_LPSM;
> @@ -1400,7 +1400,7 @@ int xhci_hub_control(struct usb_hcd *hcd, u16 typeReq, u16 wValue,
>  			 * automatically entered as on 1.0 and prior.
>  			 */
>  			if (link_state == USB_SS_PORT_LS_COMP_MOD) {
> -				if (!HCC2_CTC(xhci->hcc_params2)) {
> +				if (!(xhci->hcc_params2 & HCC2_CTC)) {
>  					xhci_dbg(xhci, "CTC flag is 0, port already supports entering compliance mode\n");
>  					break;
>  				}
> diff --git a/drivers/usb/host/xhci-mem.c b/drivers/usb/host/xhci-mem.c
> index 98abf86e0910..06ff712c9cbe 100644
> --- a/drivers/usb/host/xhci-mem.c
> +++ b/drivers/usb/host/xhci-mem.c
> @@ -463,7 +463,7 @@ struct xhci_container_ctx *xhci_alloc_container_ctx(struct xhci_hcd *xhci,
>  		return NULL;
>  
>  	ctx->type = type;
> -	ctx->size = HCC_64BYTE_CONTEXT(xhci->hcc_params) ? 2048 : 1024;
> +	ctx->size = xhci->hcc_params & HCC_64BYTE_CONTEXT ? 2048 : 1024;
>  	if (type == XHCI_CTX_TYPE_INPUT)
>  		ctx->size += CTX_SIZE(xhci->hcc_params);
>  
> @@ -1344,7 +1344,7 @@ static u32 xhci_get_endpoint_mult(struct xhci_hcd *xhci,
>  	bool lec;
>  
>  	/* xHCI 1.1 with LEC set does not use mult field, except intel eUSB2 */
> -	lec = xhci->hci_version > 0x100 && HCC2_LEC(xhci->hcc_params2);
> +	lec = xhci->hci_version > 0x100 && (xhci->hcc_params2 & HCC2_LEC);
>  
>  	/* eUSB2 double isoc bw devices are the only USB2 devices using mult */
>  	if (usb_endpoint_is_hs_isoc_double(udev, ep) &&
> @@ -1433,8 +1433,7 @@ int xhci_endpoint_init(struct xhci_hcd *xhci,
>  	ring_type = usb_endpoint_type(&ep->desc);
>  
>  	/* Ensure host supports double isoc bandwidth for eUSB2 devices */
> -	if (usb_endpoint_is_hs_isoc_double(udev, ep) &&
> -	    !HCC2_EUSB2_DIC(xhci->hcc_params2))	{
> +	if (usb_endpoint_is_hs_isoc_double(udev, ep) && !(xhci->hcc_params2 & HCC2_EUSB2_DIC))	{
>  		dev_dbg(&udev->dev, "Double Isoc Bandwidth not supported by xhci\n");
>  		return -EINVAL;
>  	}
> diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c
> index 7b7c61d79d0a..8d69f82a997c 100644
> --- a/drivers/usb/host/xhci-ring.c
> +++ b/drivers/usb/host/xhci-ring.c
> @@ -3988,7 +3988,7 @@ static int xhci_ist_in_microseconds(u32 hcs_params2)
>  {
>  	int ist = HCS_IST_VALUE(hcs_params2);
>  
> -	if (HCS_IST_UNIT(hcs_params2))
> +	if (hcs_params2 & HCS_IST_UNIT)
>  		ist *= 8;
>  	return ist;
>  }
> @@ -4157,7 +4157,7 @@ static int xhci_queue_isoc_tx(struct xhci_hcd *xhci, gfp_t mem_flags,
>  		/* use SIA as default, if frame id is used overwrite it */
>  		sia_frame_id = TRB_SIA;
>  		if (!(urb->transfer_flags & URB_ISO_ASAP) &&
> -		    HCC_CFC(xhci->hcc_params)) {
> +		    (xhci->hcc_params & HCC_CFC)) {
>  			frame_id = xhci_get_isoc_frame_id(xhci, urb, i);
>  			if (frame_id >= 0)
>  				sia_frame_id = TRB_FRAME_ID(frame_id);
> @@ -4241,7 +4241,7 @@ static int xhci_queue_isoc_tx(struct xhci_hcd *xhci, gfp_t mem_flags,
>  	}
>  
>  	/* store the next frame id */
> -	if (HCC_CFC(xhci->hcc_params))
> +	if (xhci->hcc_params & HCC_CFC)
>  		xep->next_frame_id = urb->start_frame + num_tds * urb->interval;
>  
>  	if (xhci_to_hcd(xhci)->self.bandwidth_isoc_reqs == 0) {
> @@ -4320,7 +4320,7 @@ int xhci_queue_isoc_tx_prepare(struct xhci_hcd *xhci, gfp_t mem_flags,
>  	check_interval(urb, ep_ctx);
>  
>  	/* Calculate the start frame and put it in urb->start_frame. */
> -	if (HCC_CFC(xhci->hcc_params) && !list_empty(&ep_ring->td_list)) {
> +	if ((xhci->hcc_params & HCC_CFC) && !list_empty(&ep_ring->td_list)) {
>  		if (GET_EP_CTX_STATE(ep_ctx) ==	EP_STATE_RUNNING) {
>  			urb->start_frame = xep->next_frame_id;
>  			goto skip_start_over;
> diff --git a/drivers/usb/host/xhci-trace.h b/drivers/usb/host/xhci-trace.h
> index bf13da417f8e..c6baf82912de 100644
> --- a/drivers/usb/host/xhci-trace.h
> +++ b/drivers/usb/host/xhci-trace.h
> @@ -81,7 +81,7 @@ DECLARE_EVENT_CLASS(xhci_log_ctx,
>  	),
>  	TP_fast_assign(
>  
> -		__entry->ctx_64 = HCC_64BYTE_CONTEXT(xhci->hcc_params);
> +		__entry->ctx_64 = xhci->hcc_params & HCC_64BYTE_CONTEXT;
>  		__entry->ctx_type = ctx->type;
>  		__entry->ctx_dma = ctx->dma;
>  		__entry->ctx_va = ctx->bytes;
> diff --git a/drivers/usb/host/xhci.c b/drivers/usb/host/xhci.c
> index 2015d37f863e..c7a377b34661 100644
> --- a/drivers/usb/host/xhci.c
> +++ b/drivers/usb/host/xhci.c
> @@ -5483,7 +5483,7 @@ int xhci_gen_setup(struct usb_hcd *hcd, xhci_get_quirks_t get_quirks)
>  
>  	/* Set dma_mask and coherent_dma_mask to 64-bits,
>  	 * if xHC supports 64-bit addressing */
> -	if (HCC_64BIT_ADDR(xhci->hcc_params) &&
> +	if ((xhci->hcc_params & HCC_64BIT_ADDR) &&
>  			!dma_set_mask(dev, DMA_BIT_MASK(64))) {
>  		xhci_dbg(xhci, "Enabling 64-bit DMA addresses.\n");
>  		dma_set_coherent_mask(dev, DMA_BIT_MASK(64));
> -- 
> 2.50.1
> 

-- 
Sakari Ailus

  reply	other threads:[~2025-11-11  8:51 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-11-10 15:14 [PATCH 00/13] usb: xhci: Host Controller Capability Registers rework Niklas Neronin
2025-11-10 15:14 ` [PATCH 01/13] usb: xhci: remove deprecated TODO comment Niklas Neronin
2025-11-10 15:14 ` [PATCH 02/13] usb: xhci: remove unused trace operation and argument Niklas Neronin
2025-11-10 15:14 ` [PATCH 03/13] usb: xhci: use cached HCSPARAMS1 value Niklas Neronin
2025-11-10 15:14 ` [PATCH 04/13] usb: xhci: simplify handling of Structural Parameters 1 values Niklas Neronin
2025-11-10 15:14 ` [PATCH 05/13] usb: xhci: limit number of ports to 127 Niklas Neronin
2025-11-10 15:14 ` [PATCH 06/13] usb: xhci: limit number of interrupts to 128 Niklas Neronin
2025-11-10 15:14 ` [PATCH 07/13] usb: xhci: improve xhci-caps.h comments Niklas Neronin
2025-11-10 15:14 ` [PATCH 08/13] usb: xhci: simplify Isochronous Scheduling Threshold handling Niklas Neronin
2025-11-12 16:11   ` Michal Pecio
2025-11-10 15:14 ` [PATCH 09/13] usb: xhci: simplify Max Scratchpad buffer macros Niklas Neronin
2025-11-10 15:14 ` [PATCH 10/13] usb: xhci: drop xhci-caps.h dependence on xhci-ext-caps.h Niklas Neronin
2025-11-10 15:14 ` [PATCH 11/13] usb: xhci: standardize single bit-field macros Niklas Neronin
2025-11-11  8:51   ` Sakari Ailus [this message]
2025-11-13 10:34     ` Mathias Nyman
2025-11-10 15:14 ` [PATCH 12/13] usb: xhci: standardize multi " Niklas Neronin
2025-11-11  6:10   ` kernel test robot
2025-11-11 10:00   ` kernel test robot
2025-11-10 15:14 ` [PATCH 13/13] usb: xhci: use 64-bit Addressing Capability macro Niklas Neronin

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=aRL5HQw2-a94FPpA@kekkonen.localdomain \
    --to=sakari.ailus@linux.intel.com \
    --cc=amardeep.rai@intel.com \
    --cc=linux-usb@vger.kernel.org \
    --cc=mathias.nyman@linux.intel.com \
    --cc=niklas.neronin@linux.intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox