netdev - Re: [PATCH 1/8] ixgbe: Out of line ixgbe_read/write

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:	Mon, 19 May 2014 22:00:52 +0000
From:	"Rustad, Mark D" <mark.d.rustad@...el.com>
To:	Andi Kleen <andi@...stfloor.org>,
	"Kirsher, Jeffrey T" <jeffrey.t.kirsher@...el.com>
CC:	"<linux-kernel@...r.kernel.org>" <linux-kernel@...r.kernel.org>,
	"akpm@...ux-foundation.org" <akpm@...ux-foundation.org>,
	Andi Kleen <ak@...ux.intel.com>,
	Netdev <netdev@...r.kernel.org>
Subject: Re: [PATCH 1/8] ixgbe: Out of line ixgbe_read/write_reg

On May 16, 2014, at 2:43 PM, Andi Kleen <andi@...stfloor.org> wrote:

> From: Andi Kleen <ak@...ux.intel.com>
> 
> ixgbe_read_reg and ixgbe_write_reg are frequently called and are very big
> because they have complex error handling code.

Actually, this patch doesn't do anything to ixgbe_write_reg, which would almost certainly be very bad for performance, but instead changes ixgbe_write_reg64. The latter is not in a performance-sensitive path, but is only called from one site, so there is little reason to take it out-of-line.

I already have a patch in queue to make ixgbe_read_reg out-of-line, because it does have a very costly memory footprint inline, as you have found.

> Moving them out of line saves ~27k text in the ixgbe driver.
> 
>   text	   data	    bss	    dec	    hex	filename
> 14220873	2008072	1507328	17736273	10ea251	vmlinux-before-ixgbe
> 14193673	2003976	1507328	17704977	10e2811	vmlinux-ixgbe
> 
> Cc: netdev@...r.kernel.org
> Cc: Jeff Kirsher <jeffrey.t.kirsher@...el.com>
> Signed-off-by: Andi Kleen <ak@...ux.intel.com>
> ---
> drivers/net/ethernet/intel/ixgbe/ixgbe_common.h | 22 ++--------------------
> drivers/net/ethernet/intel/ixgbe/ixgbe_main.c   | 22 ++++++++++++++++++++++
> 2 files changed, 24 insertions(+), 20 deletions(-)
> 
> diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h b/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
> index f12c40f..05f094d 100644
> --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
> +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_common.h
> @@ -162,28 +162,10 @@ static inline void writeq(u64 val, void __iomem *addr)
> }
> #endif
> 
> -static inline void ixgbe_write_reg64(struct ixgbe_hw *hw, u32 reg, u64 value)
> -{
> -	u8 __iomem *reg_addr = ACCESS_ONCE(hw->hw_addr);
> +void ixgbe_write_reg64(struct ixgbe_hw *hw, u32 reg, u64 value);
> +u32 ixgbe_read_reg(struct ixgbe_hw *hw, u32 reg);
> 
> -	if (ixgbe_removed(reg_addr))
> -		return;
> -	writeq(value, reg_addr + reg);
> -}
> #define IXGBE_WRITE_REG64(a, reg, value) ixgbe_write_reg64((a), (reg), (value))
> -
> -static inline u32 ixgbe_read_reg(struct ixgbe_hw *hw, u32 reg)
> -{
> -	u8 __iomem *reg_addr = ACCESS_ONCE(hw->hw_addr);
> -	u32 value;
> -
> -	if (ixgbe_removed(reg_addr))
> -		return IXGBE_FAILED_READ_REG;
> -	value = readl(reg_addr + reg);
> -	if (unlikely(value == IXGBE_FAILED_READ_REG))
> -		ixgbe_check_remove(hw, reg);
> -	return value;
> -}
> #define IXGBE_READ_REG(a, reg) ixgbe_read_reg((a), (reg))
> 
> #define IXGBE_WRITE_REG_ARRAY(a, reg, offset, value) \
> diff --git a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
> index d62e7a2..5f81f62 100644
> --- a/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
> +++ b/drivers/net/ethernet/intel/ixgbe/ixgbe_main.c
> @@ -371,6 +371,28 @@ void ixgbe_write_pci_cfg_word(struct ixgbe_hw *hw, u32 reg, u16 value)
> 	pci_write_config_word(adapter->pdev, reg, value);
> }
> 
> +void ixgbe_write_reg64(struct ixgbe_hw *hw, u32 reg, u64 value)
> +{
> +	u8 __iomem *reg_addr = ACCESS_ONCE(hw->hw_addr);
> +
> +	if (ixgbe_removed(reg_addr))
> +		return;
> +	writeq(value, reg_addr + reg);
> +}
> +
> +u32 ixgbe_read_reg(struct ixgbe_hw *hw, u32 reg)
> +{
> +	u8 __iomem *reg_addr = ACCESS_ONCE(hw->hw_addr);
> +	u32 value;
> +
> +	if (ixgbe_removed(reg_addr))
> +		return IXGBE_FAILED_READ_REG;
> +	value = readl(reg_addr + reg);
> +	if (unlikely(value == IXGBE_FAILED_READ_REG))
> +		ixgbe_check_remove(hw, reg);
> +	return value;
> +}
> +
> static void ixgbe_service_event_complete(struct ixgbe_adapter *adapter)
> {
> 	BUG_ON(!test_bit(__IXGBE_SERVICE_SCHED, &adapter->state));
> -- 
> 1.9.0
> 
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

-- 
Mark Rustad, Networking Division, Intel Corporation

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html