From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ryusuke Sakato Date: Wed, 10 Sep 2014 07:15:25 +0000 Subject: Re: [PATCH 0/5] arm: shmobile: Lager Add USB and SATA support Message-Id: <540FFA8D.6060602@renesas.com> List-Id: References: <1387370660-20481-1-git-send-email-valentine.barshak@cogentembedded.com> In-Reply-To: <1387370660-20481-1-git-send-email-valentine.barshak@cogentembedded.com> MIME-Version: 1.0 Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: quoted-printable To: linux-sh@vger.kernel.org Hi Hiep-san, I believe that I should unite the conditions of the performance test of "CO= NFIG_CPU_FREQ is not set" on Upstream-v3.16-rc5, and "CONFIG_CPU_FREQ=3Dy" on uptream-v3.17-rc2. > # cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor > ondemand In order that you may not be subject to the influence of cpufreq, you shoul= d change "CONFIG_CPU_FREQ is not set", or should perform the following setup. =E3=80=80echo -n performance > /sys/device/system/cpu/cpu0/cpufreq/scaling_= governor =E3=80=80echo -n performance > /sys/device/system/cpu/cpu1/cpufreq/scaling_= governor =E3=80=80echo -n performance > /sys/device/system/cpu/cpu2/cpufreq/scaling_= governor =E3=80=80echo -n performance > /sys/device/system/cpu/cpu3/cpufreq/scaling_= governor Thank you, Sakato. (2014/09/10 14:18), =E3=82=AB=E3=82=AA=E3=80=80=E3=83=9F=E3=83=B3=E3=80=80= =E3=83=92=E3=82=A7=E3=83=83=E3=83=97 wrote: > Hi Sakato-san >=20 > Thanks for your comment. >=20 > On 09/10/2014 10:21 AM, Ryusuke Sakato wrote: >> Hi Hiep-san, >> >> Can you check as CONFIG_CPU_FREQ is not set? >> >> When you perform a benchmark, you should check a /sys/devices/system/cpu= /cpu0/cpufreq/scaling_governor >> setup of your userland. > CONFIG_CPU_FREQ=3Dy on default config, > And I have just checked /sys/devices/system/cpu/cpu0/cpufreq/scaling_gove= rnor >=20 > Before writing: > # cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor > ondemand > just after writing: > #cp /tmp/file-350mb /mnt/ ; cat /sys/devices/system/cpu/cpu1/cpufreq/scal= ing_governor > ondemand >=20 >=20 > Best regards, > Cao Minh Hiep >=20 >> Thank you, >> Sakato. >> >> (2014/09/09 19:48), =E3=82=AB=E3=82=AA=E3=80=80=E3=83=9F=E3=83=B3=E3=80= =80=E3=83=92=E3=82=A7=E3=83=83=E3=83=97 wrote: >>> Hello, >>> >>> We are testing Serial-ATA driver on linux uptream-v3.17-rc2 on Lager(R8= a7790 SoC), >>> On this version Serial-ATA's Writing speed on HDD is about 42.61MB/s. >>> This writing speed|| seems good, but on Upstream-v3.16-rc5 the writing = speed was up to 66MB/s, >>> Please see it when you have time. >>> >>> Best Regards, >>> Jinso/Cao Minh Hiep. >>> >>> >>> >>> >>> >>> >>> On 12/18/2013 09:44 PM, Valentine Barshak wrote: >>>> I've decided to bundle USB and SATA patches together, since I've done = SATA on top of USB. >>>> Most of the SATA/USBHS/PCI USB host support patches for Lager are avai= lable in the respective >>>> susbsystem git trees (for 3.14) but are still missing from the Renesas= tree. >>>> >>>> So, in order to test these, the following patches from other subsystem= trees are needed: >>>> USB: >>>> https://git.kernel.org/cgit/linux/kernel/git/helgaas/pci.git/commit/?h= =3Dpci/host-rcar&id=FB178d8b2fab3f2a9f203c13ffe80cfd6e01bdf1 >>>> >>>> https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit= /?id=C8ba8115a21226fba3211085f570b128fa271e31 >>>> https://git.kernel.org/cgit/linux/kernel/git/balbi/usb.git/commit/?h= =3Dnext&id=C3e5d2985ef720cbbdc63546a5c545ac4450d96e >>>> https://git.kernel.org/cgit/linux/kernel/git/gregkh/usb.git/commit/?h= =3Dusb-next&id=103e127d1f8f985e8a662da6537ebc5e08902ee3 >>>> https://git.kernel.org/cgit/linux/kernel/git/gregkh/usb.git/commit/?h= =3Dusb-next&id=1Ae5799ef63176cc75ec10e545cb65f620a82747 >>>> >>>> SATA: >>>> https://git.kernel.org/cgit/linux/kernel/git/tj/libata.git/commit/?h= =3Dfor-3.14&id=E67adb4e669db834c1f95cbdf99bb4e9ec8455b3 >>>> >>>> The USB channel 0 (CN6) is configured as PCI USB host if USB_RENESAS_U= SBHS_UDC is disabled. >>>> Otherwise it is configured as USBHS device. >>>> Please note that the following switches have to be set correctly as we= ll: >>>> PCI USB Host: SW5 - pin1; SW6 - pin1 >>>> USBHS device: SW5 - neutral; SW6 - pin3 >>>> >>>> Valentine Barshak (5): >>>> arm: shmobile: r8a7790: Add PCI USB host clock support >>>> arm: shmobile: r8a7790: Add SATA clock >>>> arm: shmobile: lager: Add USBHS support >>>> arm: shmobile: lager: Add internal PCI support >>>> arm: shmobile: lager: Add SATA support >>>> >>>> arch/arm/mach-shmobile/board-lager.c | 203 ++++++++++++++++++++++= +++++++++++ >>>> arch/arm/mach-shmobile/clock-r8a7790.c | 25 +++- >>>> 2 files changed, 227 insertions(+), 1 deletion(-) >>>> >>> >> >=20 > . >=20