Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1754594Ab3HEQQ3 (ORCPT ); Mon, 5 Aug 2013 12:16:29 -0400 Received: from arroyo.ext.ti.com ([192.94.94.40]:39373 "EHLO arroyo.ext.ti.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752240Ab3HEQQX (ORCPT ); Mon, 5 Aug 2013 12:16:23 -0400 From: Joel Fernandes To: Tony Lindgren , Sekhar Nori , Santosh Shilimkar , Sricharan R , Rajendra Nayak , Lokesh Vutla , Matt Porter , Grant Likely , Rob Herring , Vinod Koul , Dan Williams , Mark Brown , Benoit Cousson , Russell King , Arnd Bergmann , Olof Johansson , Balaji TK , Gururaja Hebbar , Chris Ball , Jason Kridner CC: Linux OMAP List , Linux ARM Kernel List , Linux DaVinci Kernel List , Linux Kernel Mailing List , Linux MMC List , Joel Fernandes Subject: [PATCH v3 11/12] dma: edma: Keep tracking of Pending interrupts (pending_acks) Date: Mon, 5 Aug 2013 11:14:56 -0500 Message-ID: <1375719297-12871-12-git-send-email-joelf@ti.com> X-Mailer: git-send-email 1.7.9.5 In-Reply-To: <1375719297-12871-1-git-send-email-joelf@ti.com> References: <1375719297-12871-1-git-send-email-joelf@ti.com> MIME-Version: 1.0 Content-Type: text/plain Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 3504 Lines: 111 The total_processed variable was being used to keep track of when DMA is completed for a particular descriptor. This doesn't work anymore, as the interrupt for completion can come in much later than when the total_processed variable is updated to reflect that all SG entries have been issues. Introduce another variable that reflects the actual value. This will serve a purpose different from what total_processed was required to help with. Also add code that give special treatment for the first linked set interrupt as pending_acks is handled a bit differently for that case. Signed-off-by: Joel Fernandes --- drivers/dma/edma.c | 32 +++++++++++++++++++++++++++++--- 1 file changed, 29 insertions(+), 3 deletions(-) diff --git a/drivers/dma/edma.c b/drivers/dma/edma.c index 061f0cf..eca1b47 100644 --- a/drivers/dma/edma.c +++ b/drivers/dma/edma.c @@ -59,6 +59,10 @@ struct edma_desc { int pset_nr; int total_processed; int next_setup_linkset; + + int no_first_ls_ack; + int pending_acks; + struct edmacc_param pset[0]; }; @@ -147,8 +151,7 @@ static void edma_execute(struct edma_chan *echan) struct edmacc_param tmp_param; /* If either we processed all psets or we're still not started */ - if (!echan->edesc || - echan->edesc->pset_nr == echan->edesc->total_processed) { + if (!echan->edesc || !echan->edesc->pending_acks) { /* Get next vdesc */ vdesc = vchan_next_desc(&echan->vchan); if (!vdesc) { @@ -180,6 +183,8 @@ static void edma_execute(struct edma_chan *echan) edesc->total_processed += total_link_set; + edesc->pending_acks += total_link_set; + total_left = edesc->pset_nr - edesc->total_processed; total_link_set = total_left > MAX_NR_LS ? @@ -190,6 +195,8 @@ static void edma_execute(struct edma_chan *echan) where total pset_nr is strictly within MAX_NR size */ if (total_left > total_link_set) edma_enable_interrupt(echan->slot[i]); + else + edesc->no_first_ls_ack = 1; /* Setup link between linked set 0 to set 1 */ edma_link(echan->slot[i], echan->slot[i+1]); @@ -210,6 +217,9 @@ static void edma_execute(struct edma_chan *echan) } edesc->total_processed += total_link_set; + + edesc->pending_acks += total_link_set; + total_left = edesc->pset_nr - edesc->total_processed; if (total_left) @@ -267,6 +277,8 @@ static void edma_execute(struct edma_chan *echan) edesc->total_processed += total_link_set; + edesc->pending_acks += total_link_set; + slot_off = total_link_set + ls_cur_off - 1; if (edesc->total_processed == edesc->pset_nr) @@ -498,8 +510,22 @@ static void edma_callback(unsigned ch_num, u16 ch_status, void *data) edesc = echan->edesc; + if (edesc->pending_acks < MAX_NR_LS) + edesc->pending_acks = 0; + else + edesc->pending_acks -= MAX_NR_LS; + + if (edesc->no_first_ls_ack) { + if (edesc->pending_acks < MAX_NR_LS) + edesc->pending_acks = 0; + else + edesc->pending_acks -= MAX_NR_LS; + + edesc->no_first_ls_ack = 0; + } + if (edesc) { - if (edesc->total_processed == edesc->pset_nr) { + if (!edesc->pending_acks) { dev_dbg(dev, "Transfer complete," " stopping channel %d\n", ch_num); edma_stop(echan->ch_num); -- 1.7.9.5 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/