Re: [PATCH V2 net-next 1/6] ethtool: add support to set/get tx copybreak buf size via ethtool
flat view
From: huangguangbin (A) <hidden>
Date: 2021-09-29 02:20:01
Also in:
linux-s390
On 2021/9/25 7:05, Michal Kubecek wrote:
On Fri, Sep 24, 2021 at 10:29:54PM +0800, Guangbin Huang wrote:quoted
From: Hao Chen <redacted> Add support for ethtool to set/get tx copybreak buf size. Signed-off-by: Hao Chen <redacted> Signed-off-by: Guangbin Huang <redacted> --- Documentation/networking/ethtool-netlink.rst | 24 ++++++++++++++++++++ include/uapi/linux/ethtool.h | 1 + net/ethtool/common.c | 1 + net/ethtool/ioctl.c | 1 + 4 files changed, 27 insertions(+)diff --git a/Documentation/networking/ethtool-netlink.rst b/Documentation/networking/ethtool-netlink.rst index d9b55b7a1a4d..a47b0255aaf9 100644 --- a/Documentation/networking/ethtool-netlink.rst +++ b/Documentation/networking/ethtool-netlink.rst@@ -1521,6 +1521,30 @@ Kernel response contents: ``ETHTOOL_A_PHC_VCLOCKS_INDEX`` s32 PHC index array ==================================== ====== ========================== +TUNABLE_SET +=========== + +Request contents: + + ===================================== ====== ========================== + ``ETHTOOL_TX_COPYBREAK_BUF_SIZE`` u32 buf size for tx copybreak + ===================================== ====== ========================== + +Tx copybreak buf size is used for tx copybreak feature, the feature is used +for small size packet or frag. It adds a queue based tx shared bounce buffer +to memcpy the small packet when the len of xmitted skb is below tx_copybreak +(value to distinguish small size and normal size), and reduce the overhead +of dma map and unmap when IOMMU is on. + +TUNABLE_GET +=========== + +Kernel response contents: + + ==================================== ====== ========================== + ``ETHTOOL_TX_COPYBREAK_BUF_SIZE`` u32 buf size for tx copybreak + ==================================== ====== ==========================I have to repeat my concerns expressed in https://lore.kernel.org/netdev/20210826072618.2lu6spapkzdcuhyv@lion.mk-sys.cz (local) and earlier in more details in https://lore.kernel.org/netdev/20200325164958.GZ31519@unicorn.suse.cz (local) That being said, I don't understand why this patch adds description of two new message types to the documentation of ethtool netlink API but it does not actually add them to the API. Instead, it adds the new tunable to ioctl API. Michal
Hi Michal, thanks for your opinion. Is there any documentation for ioctl API? We didn't find it. Or we add a new documentation of ioctl API for the new tunable? Guangbin