From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753107AbdJLDUE (ORCPT ); Wed, 11 Oct 2017 23:20:04 -0400 Received: from mailgw02.mediatek.com ([210.61.82.184]:42378 "EHLO mailgw02.mediatek.com" rhost-flags-OK-FAIL-OK-FAIL) by vger.kernel.org with ESMTP id S1751347AbdJLDUB (ORCPT ); Wed, 11 Oct 2017 23:20:01 -0400 X-UUID: 903cbefad2114274a235406eabc98548-20171012 Message-ID: <1507778395.21840.15.camel@mtkswgap22> Subject: Re: [PATCH v3 5/9] soc: mediatek: pwrap: add pwrap_write32 for writing in 32-bit mode From: Sean Wang To: Matthias Brugger CC: , , , , , , , , , , Date: Thu, 12 Oct 2017 11:19:55 +0800 In-Reply-To: References: <236a04acb383fc655549bc345a16a2d015e5727d.1502779753.git.sean.wang@mediatek.com> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.2.3-0ubuntu6 Content-Transfer-Encoding: 7bit MIME-Version: 1.0 X-MTK: N Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Tue, 2017-10-10 at 11:38 +0200, Matthias Brugger wrote: > > On 08/15/2017 11:09 AM, sean.wang@mediatek.com wrote: > > From: Sean Wang > > > > Some regulators such as MediaTek MT6380 also has to be written in > > 32-bit mode. So the patch adds pwrap_write32, rename old pwrap_write > > into pwrap_write16 and one additional function pointer is introduced > > for increasing flexibility allowing the determination which mode is > > used by the pwrap slave detection through device tree. > > > > Signed-off-by: Chenglin Xu > > Signed-off-by: Chen Zhong > > Signed-off-by: Sean Wang > > --- > > drivers/soc/mediatek/mtk-pmic-wrap.c | 63 +++++++++++++++++++++++++++--------- > > 1 file changed, 47 insertions(+), 16 deletions(-) > > > > diff --git a/drivers/soc/mediatek/mtk-pmic-wrap.c b/drivers/soc/mediatek/mtk-pmic-wrap.c > > index 7cd581b..9d1f4c6 100644 > > --- a/drivers/soc/mediatek/mtk-pmic-wrap.c > > +++ b/drivers/soc/mediatek/mtk-pmic-wrap.c > > @@ -506,6 +506,7 @@ struct pwrap_slv_type { > > * which type is used by the detection through device tree. > > */ > > int (*pwrap_read)(struct pmic_wrapper *wrp, u32 adr, u32 *rdata); > > + int (*pwrap_write)(struct pmic_wrapper *wrp, u32 adr, u32 wdata); > > }; > > > > struct pmic_wrapper { > > @@ -600,22 +601,6 @@ static int pwrap_wait_for_state(struct pmic_wrapper *wrp, > > } while (1); > > } > > > > -static int pwrap_write(struct pmic_wrapper *wrp, u32 adr, u32 wdata) > > -{ > > - int ret; > > - > > - ret = pwrap_wait_for_state(wrp, pwrap_is_fsm_idle); > > - if (ret) { > > - pwrap_leave_fsm_vldclr(wrp); > > - return ret; > > - } > > - > > - pwrap_writel(wrp, (1 << 31) | ((adr >> 1) << 16) | wdata, > > - PWRAP_WACS2_CMD); > > - > > - return 0; > > -} > > - > > static int pwrap_read16(struct pmic_wrapper *wrp, u32 adr, u32 *rdata) > > { > > int ret; > > @@ -672,6 +657,49 @@ static int pwrap_read(struct pmic_wrapper *wrp, u32 adr, u32 *rdata) > > return wrp->slave->pwrap_read(wrp, adr, rdata); > > } > > > > +static int pwrap_write16(struct pmic_wrapper *wrp, u32 adr, u32 wdata) > > +{ > > + int ret; > > + > > + ret = pwrap_wait_for_state(wrp, pwrap_is_fsm_idle); > > + if (ret) { > > + pwrap_leave_fsm_vldclr(wrp); > > + return ret; > > + } > > + > > + pwrap_writel(wrp, (1 << 31) | ((adr >> 1) << 16) | wdata, > > + PWRAP_WACS2_CMD); > > + > > + return 0; > > +} > > + > > +static int pwrap_write32(struct pmic_wrapper *wrp, u32 adr, u32 wdata) > > +{ > > + int ret, msb, rdata; > > + > > + for (msb = 0; msb < 2; msb++) { > > + ret = pwrap_wait_for_state(wrp, pwrap_is_fsm_idle); > > + if (ret) { > > + pwrap_leave_fsm_vldclr(wrp); > > + return ret; > > + } > > + > > + pwrap_writel(wrp, (1 << 31) | (msb << 30) | (adr << 16) | > > + ((wdata >> (msb * 16)) & 0xffff), > > + PWRAP_WACS2_CMD); > > + > > + if (!msb) > > + pwrap_read(wrp, adr, &rdata); > > Just so that I understand, you have to read back the half-written register > before you can write the second part? > Yup, the pwrap_read operation is the requirement of hardware used for the synchronization between two successive 16-bit pwrap_writel operations composing one 32-bit bus writing. Otherwise, we'll find the result fails for the lower 16-bit pwrap writing. > Other then that it looks fine to me. > > Regards, > Matthias