From mboxrd@z Thu Jan 1 00:00:00 1970 From: Sinan Kaya Subject: [PATCH V4 0/3] dmaengine: qcom_hidma: add error reporting Date: Wed, 31 Aug 2016 11:10:26 -0400 Message-ID: <1472656229-12955-1-git-send-email-okaya@codeaurora.org> Return-path: Received: from smtp.codeaurora.org ([198.145.29.96]:38439 "EHLO smtp.codeaurora.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S935106AbcHaPKi (ORCPT ); Wed, 31 Aug 2016 11:10:38 -0400 Sender: linux-arm-msm-owner@vger.kernel.org List-Id: linux-arm-msm@vger.kernel.org To: dmaengine@vger.kernel.org, timur@codeaurora.org, cov@codeaurora.org, vinod.koul@intel.com, jcm@redhat.com Cc: agross@codeaurora.org, linux-arm-msm@vger.kernel.org, linux-arm-kernel@lists.infradead.org, Sinan Kaya Three changes in the error handling area: 1. There is a race condition between data transfer callback and descriptor free code. The callback routine may decide to clear the resources even though the descriptor has not yet been freed. 2. DMA Engine framework now supports direct error reporting to the client via the callback. Pass the DMA errors to the client by passing a result argument. The HW only supports a generic error when something goes wrong. That's why, using DMA_TRANS_ABORTED all the time. 3. The HIDMA driver is capable of error detection. However, the error was not being passed back to the client when tx_status API is called. ------------------------ Changes from v3 (http://www.spinics.net/lists/dmaengine/msg10742.html) ------------------------ - Clarify that the reset task is gone and also the reset behavior has changed. The driver will no longer recover the HW automatically. The driver depends on the client to call terminate_channel. Sinan Kaya (3): dmaengine: qcom_hidma: release the descriptor before the callback dmaengine: qcom_hidma: report transfer errors with new interface dmaengine: qcom_hidma: add error reporting for tx_status drivers/dma/qcom/hidma.c | 50 +++++++++++++++++++++++++++++++++++---------- drivers/dma/qcom/hidma.h | 2 +- drivers/dma/qcom/hidma_ll.c | 32 +++++++---------------------- 3 files changed, 47 insertions(+), 37 deletions(-) -- 1.9.1 From mboxrd@z Thu Jan 1 00:00:00 1970 From: okaya@codeaurora.org (Sinan Kaya) Date: Wed, 31 Aug 2016 11:10:26 -0400 Subject: [PATCH V4 0/3] dmaengine: qcom_hidma: add error reporting Message-ID: <1472656229-12955-1-git-send-email-okaya@codeaurora.org> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Three changes in the error handling area: 1. There is a race condition between data transfer callback and descriptor free code. The callback routine may decide to clear the resources even though the descriptor has not yet been freed. 2. DMA Engine framework now supports direct error reporting to the client via the callback. Pass the DMA errors to the client by passing a result argument. The HW only supports a generic error when something goes wrong. That's why, using DMA_TRANS_ABORTED all the time. 3. The HIDMA driver is capable of error detection. However, the error was not being passed back to the client when tx_status API is called. ------------------------ Changes from v3 (http://www.spinics.net/lists/dmaengine/msg10742.html) ------------------------ - Clarify that the reset task is gone and also the reset behavior has changed. The driver will no longer recover the HW automatically. The driver depends on the client to call terminate_channel. Sinan Kaya (3): dmaengine: qcom_hidma: release the descriptor before the callback dmaengine: qcom_hidma: report transfer errors with new interface dmaengine: qcom_hidma: add error reporting for tx_status drivers/dma/qcom/hidma.c | 50 +++++++++++++++++++++++++++++++++++---------- drivers/dma/qcom/hidma.h | 2 +- drivers/dma/qcom/hidma_ll.c | 32 +++++++---------------------- 3 files changed, 47 insertions(+), 37 deletions(-) -- 1.9.1