RE: [PATCH 22/29] net_dma: poll for a descriptor after allocation failure

[PATCH 00/29] ioatdma: towards raid support · Dan Williams <hidden> · 2009-09-04
[PATCH 01/29] ioat: move to drivers/dma/ioat/ · Dan Williams <hidden> · 2009-09-04
[PATCH 02/29] ioat: move definitions to dma.h · Dan Williams <hidden> · 2009-09-04
[PATCH 03/29] ioat: convert ioat_probe to pcim/devm · Dan Williams <hidden> · 2009-09-04
[PATCH 04/29] ioat: cleanup some long deref chains and 80 column collisions · Dan Williams <hidden> · 2009-09-04
[PATCH 05/29] ioat: kill function prototype ifdef guards · Dan Williams <hidden> · 2009-09-04
[PATCH 07/29] ioat: fix type mismatch for ->dmacount · Dan Williams <hidden> · 2009-09-04
[PATCH 08/29] ioat: define descriptor control bit-field · Dan Williams <hidden> · 2009-09-04
[PATCH 11/29] ioat: prepare the code for ioat[12]_dma_chan split · Dan Williams <hidden> · 2009-09-04
[PATCH 13/29] ioat1: kill unused unmap parameters · Dan Williams <hidden> · 2009-09-04
[PATCH 17/29] ioat: preserve chanctrl bits when re-arming interrupts · Dan Williams <hidden> · 2009-09-04
[PATCH 16/29] ioat: ignore reserved bits for chancnt and xfercap · Dan Williams <hidden> · 2009-09-04
[PATCH 18/29] ioat: ___devinit annotate the initialization paths · Dan Williams <hidden> · 2009-09-04
[PATCH 21/29] ioat2,3: dynamically resize descriptor ring · Dan Williams <hidden> · 2009-09-04
RE: [PATCH 21/29] ioat2,3: dynamically resize descriptor ring · Sosnowski, Maciej <hidden> · 2009-09-14
Re: [PATCH 21/29] ioat2,3: dynamically resize descriptor ring · Dan Williams <hidden> · 2009-09-15
[PATCH 20/29] ioat: switch watchdog and reset handler from workqueue to timer · Dan Williams <hidden> · 2009-09-04
RE: [PATCH 20/29] ioat: switch watchdog and reset handler from workqueue to timer · Sosnowski, Maciej <hidden> · 2009-09-14
[PATCH 24/29] fsldma: implement a private tx_list · Dan Williams <hidden> · 2009-09-04
Re: [PATCH 24/29] fsldma: implement a private tx_list · Dan Williams <hidden> · 2009-09-04
[PATCH 26/29] ioat: implement a private tx_list · Dan Williams <hidden> · 2009-09-04
RE: [PATCH 26/29] ioat: implement a private tx_list · Sosnowski, Maciej <hidden> · 2009-09-14
[PATCH 27/29] mv_xor: implement a private tx_list · Dan Williams <hidden> · 2009-09-04
[PATCH 29/29] ioat2, 3: cacheline align software descriptor allocations · Dan Williams <hidden> · 2009-09-04
RE: [PATCH 29/29] ioat2, 3: cacheline align software descriptor allocations · Sosnowski, Maciej <hidden> · 2009-09-14
[PATCH 28/29] dmaengine: kill tx_list · Dan Williams <hidden> · 2009-09-04
RE: [PATCH 28/29] dmaengine: kill tx_list · Sosnowski, Maciej <hidden> · 2009-09-14
[PATCH 25/29] iop-adma: implement a private tx_list · Dan Williams <hidden> · 2009-09-04
[PATCH 23/29] dw_dmac: implement a private tx_list · Dan Williams <hidden> · 2009-09-04
[PATCH 22/29] net_dma: poll for a descriptor after allocation failure · Dan Williams <hidden> · 2009-09-04
RE: [PATCH 22/29] net_dma: poll for a descriptor after allocation failure · Sosnowski, Maciej <hidden> · 2009-09-14
[PATCH 19/29] ioat1: trim ioat_dma_desc_sw · Dan Williams <hidden> · 2009-09-04
RE: [PATCH 19/29] ioat1: trim ioat_dma_desc_sw · Sosnowski, Maciej <hidden> · 2009-09-14
[PATCH 14/29] ioat: add some dev_dbg() calls · Dan Williams <hidden> · 2009-09-04
[PATCH 15/29] ioat: cleanup completion status reads · Dan Williams <hidden> · 2009-09-04
[PATCH 12/29] ioat2,3: convert to a true ring buffer · Dan Williams <hidden> · 2009-09-04
[PATCH 09/29] ioat1: move descriptor allocation from submit to prep · Dan Williams <hidden> · 2009-09-04
[PATCH 10/29] ioat: fix self test interrupts · Dan Williams <hidden> · 2009-09-04
[PATCH 06/29] ioat: split ioat_dma_probe into core/version-specific routines · Dan Williams <hidden> · 2009-09-04

From: Sosnowski, Maciej <hidden>
Date: 2009-09-14 15:00:52
Also in: linux-raid, lkml

Williams, Dan J wrote:

Handle descriptor allocation failures by polling for a descriptor.  The
driver will force forward progress when polled.  In the best case this
polling interval will be the time it takes for one dma memcpy
transaction to complete.  In the worst case, channel hang, we will need
to wait 100ms for the cleanup watchdog to fire (ioatdma driver).

Signed-off-by: Dan Williams <redacted>
---

Signed-off-by: Maciej Sosnowski <redacted>

`h`	back out one level
`j`	next message in thread
`k`	previous message in thread
`l`	drill in
`Esc`	close help / fold thread tree
`?`	toggle this help