From mboxrd@z Thu Jan 1 00:00:00 1970 From: =?UTF-8?B?UmFmYcWCIEtvemlr?= Subject: Re: [PATCH 1/4] igb_uio: add wc option Date: Thu, 28 Jun 2018 15:08:32 +0200 Message-ID: References: <1523455637-31719-1-git-send-email-rk@semihalf.com> <1523455637-31719-2-git-send-email-rk@semihalf.com> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Cc: dev@dpdk.org, Marcin Wojtas , =?UTF-8?Q?Micha=C5=82_Krawczyk?= , "Tzalik, Guy" , "Schmeilin, Evgeny" , "Matushevsky, Alexander" , "Chauskin, Igor" To: Ferruh Yigit Return-path: Received: from mail-lf0-f67.google.com (mail-lf0-f67.google.com [209.85.215.67]) by dpdk.org (Postfix) with ESMTP id 34C078E60 for ; Thu, 28 Jun 2018 15:08:33 +0200 (CEST) Received: by mail-lf0-f67.google.com with SMTP id g21-v6so4169430lfb.4 for ; Thu, 28 Jun 2018 06:08:33 -0700 (PDT) In-Reply-To: List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org Sender: "dev" 2018-06-27 18:34 GMT+02:00 Ferruh Yigit : > On 4/11/2018 3:07 PM, Rafal Kozik wrote: >> Write combining (wc) increase NIC performance by making better >> utilization of PCI bus, but cannot be use by all PMD. >> >> Parameter wc_activate add possibility to enable it for >> those PMD that could support it. >> >> Signed-off-by: Rafal Kozik > > Hi Rafal, > > First, sorry for delayed comment. > >> --- >> kernel/linux/igb_uio/igb_uio.c | 17 ++++++++++++++--- >> 1 file changed, 14 insertions(+), 3 deletions(-) >> >> diff --git a/kernel/linux/igb_uio/igb_uio.c b/kernel/linux/igb_uio/igb_uio.c >> index 4cae4dd..42e3b3f 100644 >> --- a/kernel/linux/igb_uio/igb_uio.c >> +++ b/kernel/linux/igb_uio/igb_uio.c >> @@ -30,6 +30,7 @@ struct rte_uio_pci_dev { >> int refcnt; >> }; >> >> +static int wc_activate; >> static char *intr_mode; >> static enum rte_intr_mode igbuio_intr_mode_preferred = RTE_INTR_MODE_MSIX; >> /* sriov sysfs */ >> @@ -375,9 +376,13 @@ igbuio_pci_setup_iomem(struct pci_dev *dev, struct uio_info *info, >> len = pci_resource_len(dev, pci_bar); >> if (addr == 0 || len == 0) >> return -1; >> - internal_addr = ioremap(addr, len); >> - if (internal_addr == NULL) >> - return -1; >> + if (wc_activate == 0) { >> + internal_addr = ioremap(addr, len); >> + if (internal_addr == NULL) >> + return -1; >> + } else { >> + internal_addr = NULL; >> + } > > Why we need this update at all? > What is the relation between internal_address and write combined? > > As far as I can see we are not using internal_addr at all, what do you observe > in write combined usage without this modification? > To get internal_addr memory need to be mapped. But as memory could not be mapped twice: with and without WC it should be skipped for WC. [1] WC does not work without this modification. [1] https://www.kernel.org/doc/ols/2008/ols2008v2-pages-135-144.pdf section 5.3 and 5.4 >> info->mem[n].name = name; >> info->mem[n].addr = addr; >> info->mem[n].internal_addr = internal_addr; >> @@ -638,6 +643,12 @@ MODULE_PARM_DESC(intr_mode, >> " " RTE_INTR_MODE_LEGACY_NAME " Use Legacy interrupt\n" >> "\n"); >> >> +module_param(wc_activate, int, 0); >> +MODULE_PARM_DESC(wc_activate, >> +"Activate support for write combining (WC) (default=0)\n" >> +" 0 - disable\n" >> +" other - enable\n"); >> + >> MODULE_DESCRIPTION("UIO driver for Intel IGB PCI cards"); >> MODULE_LICENSE("GPL"); >> MODULE_AUTHOR("Intel Corporation"); >> >