public inbox for u-boot@lists.denx.de
 help / color / mirror / Atom feed
From: Lian Minghuan-B31939 <B31939@freescale.com>
To: u-boot@lists.denx.de
Subject: [U-Boot] [PATCH] drivers/net/e1000.c: fix compile warning under 64bit mode
Date: Tue, 27 Jan 2015 11:12:40 +0800	[thread overview]
Message-ID: <54C70228.4090808@freescale.com> (raw)
In-Reply-To: <54C65C5D.7080005@freescale.com>

Hi York,

We can not use phys_addr_t and phys_size_t here.

If  CONFIG_PHYS_64BIT is defined and uboot is compiled as 32bit like 
PowerPC64 arch,
phys_addr_t and phys_size_t will be defined as 64bit, but the pointer is 
still 32bit size.
we could not convert  directly between phys_addr_t and a pointer without 
compile warning.
To avoid warning,  virt_to_phys and phys_to_virt are needed.
However, some architectures do not implemented them.

Thanks,
Minghuan


On 2015?01?26? 23:25, York Sun wrote:
>
> On 01/21/2015 11:21 PM, Minghuan Lian wrote:
>> Fix this:
>> warning: cast from pointer to integer of different size
>>
>> Signed-off-by: Minghuan Lian <Minghuan.Lian@freescale.com>
>> ---
>>   drivers/net/e1000.c | 31 +++++++++++++++++--------------
>>   1 file changed, 17 insertions(+), 14 deletions(-)
>>
>> diff --git a/drivers/net/e1000.c b/drivers/net/e1000.c
>> index 6531030..cd44222 100644
>> --- a/drivers/net/e1000.c
>> +++ b/drivers/net/e1000.c
>> @@ -4927,22 +4927,23 @@ void
>>   fill_rx(struct e1000_hw *hw)
>>   {
>>   	struct e1000_rx_desc *rd;
>> -	uint32_t flush_start, flush_end;
>> +	unsigned long flush_start, flush_end;
>>   
>>   	rx_last = rx_tail;
>>   	rd = rx_base + rx_tail;
>>   	rx_tail = (rx_tail + 1) % 8;
>>   	memset(rd, 0, 16);
>> -	rd->buffer_addr = cpu_to_le64((u32)packet);
>> +	rd->buffer_addr = cpu_to_le64((unsigned long)packet);
>>   
>>   	/*
>>   	 * Make sure there are no stale data in WB over this area, which
>>   	 * might get written into the memory while the e1000 also writes
>>   	 * into the same memory area.
>>   	 */
>> -	invalidate_dcache_range((u32)packet, (u32)packet + 4096);
>> +	invalidate_dcache_range((unsigned long)packet,
>> +				(unsigned long)packet + 4096);
>>   	/* Dump the DMA descriptor into RAM. */
>> -	flush_start = ((u32)rd) & ~(ARCH_DMA_MINALIGN - 1);
>> +	flush_start = ((unsigned long)rd) & ~(ARCH_DMA_MINALIGN - 1);
>>   	flush_end = flush_start + roundup(sizeof(*rd), ARCH_DMA_MINALIGN);
>>   	flush_dcache_range(flush_start, flush_end);
>>   
>> @@ -4963,7 +4964,7 @@ e1000_configure_tx(struct e1000_hw *hw)
>>   	unsigned long tipg, tarc;
>>   	uint32_t ipgr1, ipgr2;
>>   
>> -	E1000_WRITE_REG(hw, TDBAL, (u32) tx_base);
>> +	E1000_WRITE_REG(hw, TDBAL, (unsigned long)tx_base);
>>   	E1000_WRITE_REG(hw, TDBAH, 0);
>>   
>>   	E1000_WRITE_REG(hw, TDLEN, 128);
>> @@ -5107,7 +5108,7 @@ e1000_configure_rx(struct e1000_hw *hw)
>>   		E1000_WRITE_FLUSH(hw);
>>   	}
>>   	/* Setup the Base and Length of the Rx Descriptor Ring */
>> -	E1000_WRITE_REG(hw, RDBAL, (u32) rx_base);
>> +	E1000_WRITE_REG(hw, RDBAL, (unsigned long)rx_base);
>>   	E1000_WRITE_REG(hw, RDBAH, 0);
>>   
>>   	E1000_WRITE_REG(hw, RDLEN, 128);
>> @@ -5138,14 +5139,14 @@ e1000_poll(struct eth_device *nic)
>>   {
>>   	struct e1000_hw *hw = nic->priv;
>>   	struct e1000_rx_desc *rd;
>> -	uint32_t inval_start, inval_end;
>> +	unsigned long inval_start, inval_end;
>>   	uint32_t len;
>>   
>>   	/* return true if there's an ethernet packet ready to read */
>>   	rd = rx_base + rx_last;
>>   
>>   	/* Re-load the descriptor from RAM. */
>> -	inval_start = ((u32)rd) & ~(ARCH_DMA_MINALIGN - 1);
>> +	inval_start = ((unsigned long)rd) & ~(ARCH_DMA_MINALIGN - 1);
>>   	inval_end = inval_start + roundup(sizeof(*rd), ARCH_DMA_MINALIGN);
>>   	invalidate_dcache_range(inval_start, inval_end);
>>   
>> @@ -5154,8 +5155,9 @@ e1000_poll(struct eth_device *nic)
>>   	/*DEBUGOUT("recv: packet len=%d \n", rd->length); */
>>   	/* Packet received, make sure the data are re-loaded from RAM. */
>>   	len = le32_to_cpu(rd->length);
>> -	invalidate_dcache_range((u32)packet,
>> -				(u32)packet + roundup(len, ARCH_DMA_MINALIGN));
>> +	invalidate_dcache_range((unsigned long)packet,
>> +				(unsigned long)packet +
>> +				roundup(len, ARCH_DMA_MINALIGN));
>>   	NetReceive((uchar *)packet, len);
>>   	fill_rx(hw);
>>   	return 1;
>> @@ -5170,7 +5172,7 @@ static int e1000_transmit(struct eth_device *nic, void *txpacket, int length)
>>   	struct e1000_hw *hw = nic->priv;
>>   	struct e1000_tx_desc *txp;
>>   	int i = 0;
>> -	uint32_t flush_start, flush_end;
>> +	unsigned long flush_start, flush_end;
>>   
>>   	txp = tx_base + tx_tail;
>>   	tx_tail = (tx_tail + 1) % 8;
>> @@ -5180,10 +5182,11 @@ static int e1000_transmit(struct eth_device *nic, void *txpacket, int length)
>>   	txp->upper.data = 0;
>>   
>>   	/* Dump the packet into RAM so e1000 can pick them. */
>> -	flush_dcache_range((u32)nv_packet,
>> -			   (u32)nv_packet + roundup(length, ARCH_DMA_MINALIGN));
>> +	flush_dcache_range((unsigned long)nv_packet,
>> +			   (unsigned long)nv_packet +
>> +			   roundup(length, ARCH_DMA_MINALIGN));
>>   	/* Dump the descriptor into RAM as well. */
>> -	flush_start = ((u32)txp) & ~(ARCH_DMA_MINALIGN - 1);
>> +	flush_start = ((unsigned long)txp) & ~(ARCH_DMA_MINALIGN - 1);
>>   	flush_end = flush_start + roundup(sizeof(*txp), ARCH_DMA_MINALIGN);
>>   	flush_dcache_range(flush_start, flush_end);
>>   
>>
> Please consider phys_addr_t and phys_size_t.
>
> York

  reply	other threads:[~2015-01-27  3:12 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-01-22  5:21 [U-Boot] [PATCH] drivers/net/e1000.c: fix compile warning under 64bit mode Minghuan Lian
2015-01-22  5:21 ` [U-Boot] [PATCH 1/2] drivers/pci/pci_rom.c: " Minghuan Lian
2015-02-02 18:59   ` [U-Boot] [U-Boot, " Tom Rini
2015-01-26 15:25 ` [U-Boot] [PATCH] drivers/net/e1000.c: " York Sun
2015-01-27  3:12   ` Lian Minghuan-B31939 [this message]
2015-01-27  3:16     ` York Sun
2015-02-02 18:59 ` [U-Boot] " Tom Rini

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=54C70228.4090808@freescale.com \
    --to=b31939@freescale.com \
    --cc=u-boot@lists.denx.de \
    /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