lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date:   Fri, 27 Jan 2017 15:56:47 +0100
From:   Nicolas Ferre <nicolas.ferre@...el.com>
To:     Cristian Birsan <cristian.birsan@...rochip.com>,
        <balbi@...nel.org>, <gregkh@...uxfoundation.org>,
        <linux-arm-kernel@...ts.infradead.org>,
        <linux-usb@...r.kernel.org>, <linux-kernel@...r.kernel.org>
CC:     <ludovic.desroches@...el.com>,
        <alexandre.belloni@...e-electrons.com>,
        <boris.brezillon@...e-electrons.com>
Subject: Re: [PATCH linux-next v2 1/1] usb: gadget: udc: atmel: Update
 endpoint allocation scheme

Le 27/01/2017 à 14:47, Cristian Birsan a écrit :
> I will send a fixup patch with updates based on the comments received from Nicolas.
> 
> On 01/26/2017 05:02 PM, Nicolas Ferre wrote:
>> Le 23/01/2017 à 15:45, cristian.birsan@...rochip.com a écrit :
>>> From: Cristian Birsan <cristian.birsan@...rochip.com>
>>>
>>> Update atmel udc driver with a new enpoint allocation scheme. The data
>>> sheet requires that all endpoints are allocated in order.
>>
>> Pieces of information from the cover letter could have been added here.
>>
> 
> Ack. I'll remove the cover letter.
> 
>>> Signed-off-by: Cristian Birsan <cristian.birsan@...rochip.com>
>>> ---
>>>  drivers/usb/gadget/udc/Kconfig          |  14 ++
>>>  drivers/usb/gadget/udc/atmel_usba_udc.c | 236 +++++++++++++++++++++++++++-----
>>>  drivers/usb/gadget/udc/atmel_usba_udc.h |  10 +-
>>>  3 files changed, 227 insertions(+), 33 deletions(-)
>>
>> It can be good to run ./scripts/checkpatch.pl --strict as well. It would
>> have told you about some of the alignment and braces issues. I ran it as
>> some the code has looked weird to me once or twice...
>>
>>
> 
> Ack.
> 
>>> diff --git a/drivers/usb/gadget/udc/Kconfig b/drivers/usb/gadget/udc/Kconfig
>>> index 658b8da..4b69f28 100644
>>> --- a/drivers/usb/gadget/udc/Kconfig
>>> +++ b/drivers/usb/gadget/udc/Kconfig
>>> @@ -60,6 +60,20 @@ config USB_ATMEL_USBA
>>>  	  USBA is the integrated high-speed USB Device controller on
>>>  	  the AT32AP700x, some AT91SAM9 and AT91CAP9 processors from Atmel.
>>>  
>>> +	  The fifo_mode parameter is used to select endpoint allocation mode.
>>> +	  fifo_mode = 0 is used to let the driver autoconfigure the endpoints.
>>> +	  In this case 2 banks are allocated for isochronous endpoints and
>>
>> You mean that 2 banks can do isochronous xfers if needed, but they can
>> do other types as well, right? So maybe rephrase the sentence.
>>
> 
> For fifo_mode = 0 (autoconfigure case) the driver allocates 2 banks only for
> isochronous endpoints as required by the reference manual. For other endpoints
> it will allocate only 1 bank because it cannot know in advance the number of
> endpoints and the size of them, especially if the configfs is used to create
> the gadget. The idea is to support as many endpoints as we can in this mode.

You know that I know this. It was just a remark about the wording of
your sentence. "are allocated for isochronous" can be turned into "are
allocated so that they could be use as isochronous endpoints". Not a big
deal though.


>>> +	  only one bank is allocated for the rest of the endpoints.
>>> +
>>> +	  fifo_mode = 1 is a generic maximum fifo size (1024 bytes) configuration
>>> +	  allowing the usage of ep1 - ep6
>>> +
>>> +	  fifo_mode = 2 is a generic performance maximum fifo size (1024 bytes)
>>> +	  configuration allowing the usage of ep1 - ep3
>>> +
>>> +	  fifo_mode = 3 is a balanced performance configuration allowing the
>>> +	  the usage of ep1 - ep8
>>> +
>>>  config USB_BCM63XX_UDC
>>>  	tristate "Broadcom BCM63xx Peripheral Controller"
>>>  	depends on BCM63XX
>>> diff --git a/drivers/usb/gadget/udc/atmel_usba_udc.c b/drivers/usb/gadget/udc/atmel_usba_udc.c
>>> index 12c7687..11bbce2 100644
>>> --- a/drivers/usb/gadget/udc/atmel_usba_udc.c
>>> +++ b/drivers/usb/gadget/udc/atmel_usba_udc.c
>>> @@ -20,6 +20,7 @@
>>>  #include <linux/mfd/syscon.h>
>>>  #include <linux/platform_device.h>
>>>  #include <linux/regmap.h>
>>> +#include <linux/ctype.h>
>>>  #include <linux/usb/ch9.h>
>>>  #include <linux/usb/gadget.h>
>>>  #include <linux/usb/atmel_usba_udc.h>
>>> @@ -318,6 +319,91 @@ static inline void usba_cleanup_debugfs(struct usba_udc *udc)
>>>  }
>>>  #endif
>>>  
>>> +static ushort fifo_mode;
>>> +
>>> +/* "modprobe ... fifo_mode=1" etc */
>>
>> No need for this comment.
>>
>  
> Ack. I'll remove it.
> 
>>> +module_param(fifo_mode, ushort, 0x0);
>>> +MODULE_PARM_DESC(fifo_mode, "Endpoint configuration mode");
>>> +
>>> +/* mode 0 - uses autoconfig */
>>> +
>>> +/* mode 1 - fits in 8KB, generic max fifo configuration */
>>> +static struct usba_fifo_cfg mode_1_cfg[] = {
>>> +{ .hw_ep_num = 0, .fifo_size = 64,	.nr_banks = 1, },
>>> +{ .hw_ep_num = 1, .fifo_size = 1024,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 2, .fifo_size = 1024,	.nr_banks = 1, },
>>> +{ .hw_ep_num = 3, .fifo_size = 1024,	.nr_banks = 1, },
>>> +{ .hw_ep_num = 4, .fifo_size = 1024,	.nr_banks = 1, },
>>> +{ .hw_ep_num = 5, .fifo_size = 1024,	.nr_banks = 1, },
>>> +{ .hw_ep_num = 6, .fifo_size = 1024,	.nr_banks = 1, },
>>> +};
>>> +
>>> +/* mode 2 - fits in 8KB, performance max fifo configuration */
>>> +static struct usba_fifo_cfg mode_2_cfg[] = {
>>> +{ .hw_ep_num = 0, .fifo_size = 64,	.nr_banks = 1, },
>>> +{ .hw_ep_num = 1, .fifo_size = 1024,	.nr_banks = 3, },
>>> +{ .hw_ep_num = 2, .fifo_size = 1024,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 3, .fifo_size = 1024,	.nr_banks = 2, },
>>> +};
>>> +
>>> +/* mode 3 - fits in 8KB, mixed fifo configuration */
>>> +static struct usba_fifo_cfg mode_3_cfg[] = {
>>> +{ .hw_ep_num = 0, .fifo_size = 64,	.nr_banks = 1, },
>>> +{ .hw_ep_num = 1, .fifo_size = 1024,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 2, .fifo_size = 512,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 3, .fifo_size = 512,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 4, .fifo_size = 512,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 5, .fifo_size = 512,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 6, .fifo_size = 512,	.nr_banks = 2, },
>>> +};
>>> +
>>> +/* mode 4 - fits in 8KB, custom fifo configuration */
>>> +static struct usba_fifo_cfg mode_4_cfg[] = {
>>> +{ .hw_ep_num = 0, .fifo_size = 64,	.nr_banks = 1, },
>>> +{ .hw_ep_num = 1, .fifo_size = 512,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 2, .fifo_size = 512,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 3, .fifo_size = 8,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 4, .fifo_size = 512,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 5, .fifo_size = 512,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 6, .fifo_size = 16,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 7, .fifo_size = 8,	.nr_banks = 2, },
>>> +{ .hw_ep_num = 8, .fifo_size = 8,	.nr_banks = 2, },
>>> +};
>>> +/* Add additional configurations here */
>>> +
>>> +int usba_config_fifo_table(struct usba_udc *udc)
>>
>> Isn't is static?
>>
> 
> Ack. I'm going to make it static.
> 
>>> +{
>>> +	int n;
>>> +
>>> +	switch (fifo_mode) {
>>> +	default:
>>> +		fifo_mode = 0;
>>> +	case 0:
>>> +		udc->fifo_cfg = NULL;
>>> +		n = 0;
>>> +		break;
>>> +	case 1:
>>> +		udc->fifo_cfg = mode_1_cfg;
>>> +		n = ARRAY_SIZE(mode_1_cfg);
>>> +		break;
>>> +	case 2:
>>> +		udc->fifo_cfg = mode_2_cfg;
>>> +		n = ARRAY_SIZE(mode_2_cfg);
>>> +		break;
>>> +	case 3:
>>> +		udc->fifo_cfg = mode_3_cfg;
>>> +		n = ARRAY_SIZE(mode_3_cfg);
>>> +		break;
>>> +	case 4:
>>> +		udc->fifo_cfg = mode_4_cfg;
>>> +		n = ARRAY_SIZE(mode_4_cfg);
>>> +		break;
>>> +	}
>>> +	DBG(DBG_HW, "Setup fifo_mode %d\n", fifo_mode);
>>> +
>>> +	return n;
>>> +}
>>> +
>>>  static inline u32 usba_int_enb_get(struct usba_udc *udc)
>>>  {
>>>  	return udc->int_enb_cache;
>>> @@ -543,24 +629,17 @@ usba_ep_enable(struct usb_ep *_ep, const struct usb_endpoint_descriptor *desc)
>>>  	ep->is_isoc = 0;
>>>  	ep->is_in = 0;
>>>  
>>> -	if (maxpacket <= 8)
>>> -		ept_cfg = USBA_BF(EPT_SIZE, USBA_EPT_SIZE_8);
>>> -	else
>>> -		/* LSB is bit 1, not 0 */
>>> -		ept_cfg = USBA_BF(EPT_SIZE, fls(maxpacket - 1) - 3);
>>> -
>>> -	DBG(DBG_HW, "%s: EPT_SIZE = %lu (maxpacket = %lu)\n",
>>> +	DBG(DBG_ERR, "%s: EPT_CFG = 0x%lx (maxpacket = %lu)\n",
>>>  			ep->ep.name, ept_cfg, maxpacket);
>>>  
>>>  	if (usb_endpoint_dir_in(desc)) {
>>>  		ep->is_in = 1;
>>> -		ept_cfg |= USBA_EPT_DIR_IN;
>>> +		ep->ept_cfg |= USBA_EPT_DIR_IN;
>>>  	}
>>>  
>>>  	switch (usb_endpoint_type(desc)) {
>>>  	case USB_ENDPOINT_XFER_CONTROL:
>>> -		ept_cfg |= USBA_BF(EPT_TYPE, USBA_EPT_TYPE_CONTROL);
>>> -		ept_cfg |= USBA_BF(BK_NUMBER, USBA_BK_NUMBER_ONE);
>>> +		ep->ept_cfg |= USBA_BF(EPT_TYPE, USBA_EPT_TYPE_CONTROL);
>>>  		break;
>>>  	case USB_ENDPOINT_XFER_ISOC:
>>>  		if (!ep->can_isoc) {
>>> @@ -578,24 +657,15 @@ usba_ep_enable(struct usb_ep *_ep, const struct usb_endpoint_descriptor *desc)
>>>  			return -EINVAL;
>>>  
>>>  		ep->is_isoc = 1;
>>> -		ept_cfg |= USBA_BF(EPT_TYPE, USBA_EPT_TYPE_ISO);
>>> +		ep->ept_cfg |= USBA_BF(EPT_TYPE, USBA_EPT_TYPE_ISO);
>>> +		ep->ept_cfg |= USBA_BF(NB_TRANS, nr_trans);
>>>  
>>> -		/*
>>> -		 * Do triple-buffering on high-bandwidth iso endpoints.
>>> -		 */
>>> -		if (nr_trans > 1 && ep->nr_banks == 3)
>>> -			ept_cfg |= USBA_BF(BK_NUMBER, USBA_BK_NUMBER_TRIPLE);
>>> -		else
>>> -			ept_cfg |= USBA_BF(BK_NUMBER, USBA_BK_NUMBER_DOUBLE);
>>> -		ept_cfg |= USBA_BF(NB_TRANS, nr_trans);
>>>  		break;
>>>  	case USB_ENDPOINT_XFER_BULK:
>>> -		ept_cfg |= USBA_BF(EPT_TYPE, USBA_EPT_TYPE_BULK);
>>> -		ept_cfg |= USBA_BF(BK_NUMBER, USBA_BK_NUMBER_DOUBLE);
>>> +		ep->ept_cfg |= USBA_BF(EPT_TYPE, USBA_EPT_TYPE_BULK);
>>>  		break;
>>>  	case USB_ENDPOINT_XFER_INT:
>>> -		ept_cfg |= USBA_BF(EPT_TYPE, USBA_EPT_TYPE_INT);
>>> -		ept_cfg |= USBA_BF(BK_NUMBER, USBA_BK_NUMBER_DOUBLE);
>>> +		ep->ept_cfg |= USBA_BF(EPT_TYPE, USBA_EPT_TYPE_INT);
>>>  		break;
>>>  	}
>>>  
>>> @@ -604,7 +674,7 @@ usba_ep_enable(struct usb_ep *_ep, const struct usb_endpoint_descriptor *desc)
>>>  	ep->ep.desc = desc;
>>>  	ep->ep.maxpacket = maxpacket;
>>>  
>>> -	usba_ep_writel(ep, CFG, ept_cfg);
>>> +	usba_ep_writel(ep, CFG, ep->ept_cfg);
>>>  	usba_ep_writel(ep, CTL_ENB, USBA_EPT_ENABLE);
>>>  
>>>  	if (ep->can_dma) {
>>> @@ -1006,12 +1076,81 @@ static int atmel_usba_start(struct usb_gadget *gadget,
>>>  		struct usb_gadget_driver *driver);
>>>  static int atmel_usba_stop(struct usb_gadget *gadget);
>>>  
>>> +static struct usb_ep *atmel_usba_match_ep(
>>
>> I find the indentation of this funciton wreird...
>>
>>> +		struct usb_gadget		*gadget,
>>
>> No need for a tab here and align with open parenthesis,
>>
> 
> Ack.
> 
>>> +		struct usb_endpoint_descriptor	*desc,
>>> +		struct usb_ss_ep_comp_descriptor *ep_comp
>>> +)
>>
>> This parenthesis seems weird to me: put it after the last parameter.
>>
> 
> Ack.
>  
>>> +{
>>> +	struct usb_ep	*_ep;
>>> +	struct usba_ep *ep;
>>> +
>>> +	/* Look at endpoints until an unclaimed one looks usable */
>>> +	list_for_each_entry(_ep, &gadget->ep_list, ep_list) {
>>> +		if (usb_gadget_ep_match_desc(gadget, _ep, desc, ep_comp))
>>> +			goto found_ep;
>>> +	}
>>> +	/* Fail */
>>> +	return NULL;
>>> +
>>> +found_ep:
>>> +
>>> +	if (fifo_mode == 0) {
>>> +		/* Optimize hw fifo size based on ep type and other info */
>>> +		ep = to_usba_ep(_ep);
>>> +
>>> +		switch (usb_endpoint_type(desc)) {
>>> +
>>> +		case USB_ENDPOINT_XFER_CONTROL:
>>> +			break;
>>> +
>>> +		case USB_ENDPOINT_XFER_ISOC:
>>> +			ep->fifo_size = 1024;
>>> +			ep->nr_banks = 2;
>>> +			break;
>>> +
>>> +		case USB_ENDPOINT_XFER_BULK:
>>> +			ep->fifo_size = 512;
>>> +			ep->nr_banks = 1;
>>> +			break;
>>> +
>>> +		case USB_ENDPOINT_XFER_INT:
>>> +			if (desc->wMaxPacketSize == 0)
>>> +				ep->fifo_size =
>>> +				    roundup_pow_of_two(_ep->maxpacket_limit);
>>> +			else
>>> +				ep->fifo_size =
>>> +				    roundup_pow_of_two(le16_to_cpu(desc->wMaxPacketSize));
>>> +			ep->nr_banks = 1;
>>> +			break;
>>> +		}
>>> +
>>> +		/* It might be a little bit late to set this */
>>
>> Is it too late or not? This comment is frightening... We may need some
>> feedback from the USB guys here...
>>
> 
> If someone from USB can comment a bit on this topic I will be grateful.
> 
> The idea here is to optimize the endpoint size used by usb gadgets even if they are created
> at runtime using configfs. Calling usb_ep_set_maxpacket_limit() in probe is too early for
> the size of the endpoint used by the above gadget function to be known. Setting this
> value here it might be a bit late only if the above layer is already using internally the
> maxpacket value configured in probe. From my tests on CDC ACM, RNDIS and HID this is not
> a problem. I added the message as a warning.

Ok, indeed can be good to have feedback from Felipe or Greg.

>>> +		usb_ep_set_maxpacket_limit(&ep->ep, ep->fifo_size);
>>> +
>>> +		/* Generate ept_cfg basd on FIFO size and number of banks */
>>> +		if (ep->fifo_size  <= 8)
>>> +			ep->ept_cfg = USBA_BF(EPT_SIZE, USBA_EPT_SIZE_8);
>>> +		else
>>> +			/* LSB is bit 1, not 0 */
>>> +			ep->ept_cfg =
>>> +				USBA_BF(EPT_SIZE, fls(ep->fifo_size - 1) - 3);
>>> +
>>> +		ep->ept_cfg |= USBA_BF(BK_NUMBER, ep->nr_banks);
>>> +		ep->udc->configured_ep++;
>>
>> is it more than a boolean value?
>>
> 
> Yes. It is more than a boolean value. It counts the number of configured
> endpoints by atmel_usba_match_ep() function that is used as autoconfigure
> mechanism.
> 
>>> +	}
>>> +
>>> +return _ep;
>>
>> Indentation issue here ^
>>
> 
> Ack.
> 
>>> +}
>>> +
>>>  static const struct usb_gadget_ops usba_udc_ops = {
>>>  	.get_frame		= usba_udc_get_frame,
>>>  	.wakeup			= usba_udc_wakeup,
>>>  	.set_selfpowered	= usba_udc_set_selfpowered,
>>>  	.udc_start		= atmel_usba_start,
>>>  	.udc_stop		= atmel_usba_stop,
>>> +	.match_ep		= atmel_usba_match_ep,
>>>  };
>>>  
>>>  static struct usb_endpoint_descriptor usba_ep0_desc = {
>>> @@ -1678,7 +1817,8 @@ static irqreturn_t usba_udc_irq(int irq, void *devid)
>>>  	}
>>>  
>>>  	if (status & USBA_END_OF_RESET) {
>>> -		struct usba_ep *ep0;
>>> +		struct usba_ep *ep0, *ep;
>>> +		int i, n;
>>>  
>>>  		usba_writel(udc, INT_CLR, USBA_END_OF_RESET);
>>>  		generate_bias_pulse(udc);
>>> @@ -1717,6 +1857,16 @@ static irqreturn_t usba_udc_irq(int irq, void *devid)
>>>  		if (!(usba_ep_readl(ep0, CFG) & USBA_EPT_MAPPED))
>>>  			dev_dbg(&udc->pdev->dev,
>>>  				 "ODD: EP0 configuration is invalid!\n");
>>> +
>>> +		/* Preallocate other endpoints */
>>> +		n = fifo_mode ? udc->num_ep : udc->configured_ep;
>>> +		for (i = 1; i < n; i++) {
>>> +			ep = &udc->usba_ep[i];
>>> +			usba_ep_writel(ep, CFG, ep->ept_cfg);
>>> +			if (!(usba_ep_readl(ep, CFG) & USBA_EPT_MAPPED))
>>> +				dev_dbg(&udc->pdev->dev,
>>> +				 "ODD: EP%d configuration is invalid!\n", i);
>>
>> I know that it's the case for the EP0 just above and that we do it like
>> this in the current driver but maybe we can do more than just having a
>> debug message.
>>
> 
> As you said the debug message follows the pattern that was already in the driver.
> I never stumbled upon this corner case but for now I would like to keep the print
> here to avoid any silent failure. If the USBA_EPT_MAPPED flag is not raised by
> hardware it means that internal memory allocation in the controller went wrong (most
> of the time due to not enough memory, caused by too many / large endpoints). It's
> hard to imagine a solution that can fix this situation without reinitializing the
> usba controller. This can be fixed in a later patch if it's needed.

dev_dbg is pretty much silent all the time. dev_warn() can certainly be
considered. Then in addition to this warning, a return value that could
signal the issue on upper layers is certainly a good idea. But as you
said, certainly the topic of another patch/patch series.

> 
>>> +		}
>>>  	}
>>>  
>>>  	spin_unlock(&udc->lock);
>>> @@ -1864,6 +2014,9 @@ static int atmel_usba_stop(struct usb_gadget *gadget)
>>>  	if (gpio_is_valid(udc->vbus_pin))
>>>  		disable_irq(gpio_to_irq(udc->vbus_pin));
>>>  
>>> +	if (fifo_mode == 0)
>>> +		udc->configured_ep = 1;
>>> +
>>>  	usba_stop(udc);
>>>  
>>>  	udc->driver = NULL;
>>> @@ -1931,9 +2084,13 @@ static struct usba_ep * atmel_udc_of_init(struct platform_device *pdev,
>>>  						&flags);
>>>  	udc->vbus_pin_inverted = (flags & OF_GPIO_ACTIVE_LOW) ? 1 : 0;
>>>  
>>> -	pp = NULL;
>>> -	while ((pp = of_get_next_child(np, pp)))
>>> -		udc->num_ep++;
>>> +	if (fifo_mode == 0) {
>>> +		pp = NULL;
>>> +		while ((pp = of_get_next_child(np, pp)))
>>> +			udc->num_ep++;
>>> +		udc->configured_ep = 1;
>>> +	} else
>>
>> Braces here
>>
> 
> Ack. Will fix.
> 
>>> +		udc->num_ep = usba_config_fifo_table(udc);
>>>  
>>>  	eps = devm_kzalloc(&pdev->dev, sizeof(struct usba_ep) * udc->num_ep,
>>>  			   GFP_KERNEL);
>>> @@ -1946,7 +2103,7 @@ static struct usba_ep * atmel_udc_of_init(struct platform_device *pdev,
>>>  
>>>  	pp = NULL;
>>>  	i = 0;
>>> -	while ((pp = of_get_next_child(np, pp))) {
>>> +	while ((pp = of_get_next_child(np, pp)) && i < udc->num_ep) {
>>>  		ep = &eps[i];
>>>  
>>>  		ret = of_property_read_u32(pp, "reg", &val);
>>> @@ -1954,21 +2111,21 @@ static struct usba_ep * atmel_udc_of_init(struct platform_device *pdev,
>>>  			dev_err(&pdev->dev, "of_probe: reg error(%d)\n", ret);
>>>  			goto err;
>>>  		}
>>> -		ep->index = val;
>>> +		ep->index = fifo_mode ? udc->fifo_cfg[i].hw_ep_num : val;
>>>  
>>>  		ret = of_property_read_u32(pp, "atmel,fifo-size", &val);
>>>  		if (ret) {
>>>  			dev_err(&pdev->dev, "of_probe: fifo-size error(%d)\n", ret);
>>>  			goto err;
>>>  		}
>>> -		ep->fifo_size = val;
>>> +		ep->fifo_size = fifo_mode ? udc->fifo_cfg[i].fifo_size : val;
>>
>> No, this is where I would like to see that the value from the table is
>> still validated agains the "max value" that is stored in HW description/DT.
>> I know that all of our product are able to handle 1024 fifo size, but
>> it's more a savfe check...
>>
> 
> Ack. I'll validate against "max value".
>  
>>>  
>>>  		ret = of_property_read_u32(pp, "atmel,nb-banks", &val);
>>>  		if (ret) {
>>>  			dev_err(&pdev->dev, "of_probe: nb-banks error(%d)\n", ret);
>>>  			goto err;
>>>  		}
>>> -		ep->nr_banks = val;
>>> +		ep->nr_banks = fifo_mode ? udc->fifo_cfg[i].nr_banks : val;
>>
>> Here however, it is pretty important to check this value. For instance,
>> notice that the at91sam9x5 cannot fullfil the "mode 2" configuration: we
>> can't silentely ignore the HW specifications given by DT.
>>
> 
> Ack. Good catch. Although at91sam9x5 has a slightly different name for the usb
> device controller in the reference manual, it is basically the same ip. 
> 
>>>  		ep->can_dma = of_property_read_bool(pp, "atmel,can-dma");
>>>  		ep->can_isoc = of_property_read_bool(pp, "atmel,can-isoc");
>>> @@ -2000,6 +2157,21 @@ static struct usba_ep * atmel_udc_of_init(struct platform_device *pdev,
>>>  		ep->ep.caps.dir_in = true;
>>>  		ep->ep.caps.dir_out = true;
>>>  
>>> +		if (fifo_mode != 0) {
>>> +			/*
>>> +			 * Generate ept_cfg based on FIFO size and
>>> +			 * banks number
>>> +			 */
>>> +			if (ep->fifo_size  <= 8)
>>> +				ep->ept_cfg = USBA_BF(EPT_SIZE, USBA_EPT_SIZE_8);
>>> +			else
>>> +				/* LSB is bit 1, not 0 */
>>> +				ep->ept_cfg =
>>> +				  USBA_BF(EPT_SIZE, fls(ep->fifo_size - 1) - 3);
>>> +
>>> +			ep->ept_cfg |= USBA_BF(BK_NUMBER, ep->nr_banks);
>>> +		}
>>> +
>>>  		if (i)
>>>  			list_add_tail(&ep->ep.ep_list, &udc->gadget.ep_list);
>>>  
>>> diff --git a/drivers/usb/gadget/udc/atmel_usba_udc.h b/drivers/usb/gadget/udc/atmel_usba_udc.h
>>> index b03b2eb..9551b70 100644
>>> --- a/drivers/usb/gadget/udc/atmel_usba_udc.h
>>> +++ b/drivers/usb/gadget/udc/atmel_usba_udc.h
>>> @@ -275,6 +275,12 @@ struct usba_dma_desc {
>>>  	u32 ctrl;
>>>  };
>>>  
>>> +struct usba_fifo_cfg {
>>> +	u8			hw_ep_num;
>>> +	u16			fifo_size;
>>> +	u8			nr_banks;
>>> +};
>>> +
>>>  struct usba_ep {
>>>  	int					state;
>>>  	void __iomem				*ep_regs;
>>> @@ -293,7 +299,7 @@ struct usba_ep {
>>>  	unsigned int				can_isoc:1;
>>>  	unsigned int				is_isoc:1;
>>>  	unsigned int				is_in:1;
>>> -
>>> +	unsigned long				ept_cfg;
>>>  #ifdef CONFIG_USB_GADGET_DEBUG_FS
>>>  	u32					last_dma_status;
>>>  	struct dentry				*debugfs_dir;
>>> @@ -338,6 +344,8 @@ struct usba_udc {
>>>  	int vbus_pin;
>>>  	int vbus_pin_inverted;
>>>  	int num_ep;
>>> +	int configured_ep;
>>> +	struct usba_fifo_cfg *fifo_cfg;
>>>  	struct clk *pclk;
>>>  	struct clk *hclk;
>>>  	struct usba_ep *usba_ep;

Regards,

-- 
Nicolas Ferre

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ