From: "Bryan O'Sullivan" <bos@pathscale.com>
To: akpm@osdl.org, rdreier@cisco.com, mst@mellanox.co.il
Cc: openib-general@openib.org, linux-kernel@vger.kernel.org,
netdev@vger.kernel.org
Subject: [PATCH 17 of 39] IB/ipath - use more appropriate gfp flags
Date: Thu, 29 Jun 2006 14:41:08 -0700 [thread overview]
Message-ID: <9d943b828776136a2bb7.1151617268@eng-12.pathscale.com> (raw)
In-Reply-To: <patchbomb.1151617251@eng-12.pathscale.com>
This helps us to survive better when memory is fragmented.
Signed-off-by: Dave Olson <dave.olson@qlogic.com>
Signed-off-by: Bryan O'Sullivan <bryan.osullivan@qlogic.com>
diff -r fd5e733f02ac -r 9d943b828776 drivers/infiniband/hw/ipath/ipath_file_ops.c
--- a/drivers/infiniband/hw/ipath/ipath_file_ops.c Thu Jun 29 14:33:25 2006 -0700
+++ b/drivers/infiniband/hw/ipath/ipath_file_ops.c Thu Jun 29 14:33:25 2006 -0700
@@ -705,6 +705,15 @@ static int ipath_create_user_egr(struct
unsigned e, egrcnt, alloced, egrperchunk, chunk, egrsize, egroff;
size_t size;
int ret;
+ gfp_t gfp_flags;
+
+ /*
+ * GFP_USER, but without GFP_FS, so buffer cache can be
+ * coalesced (we hope); otherwise, even at order 4,
+ * heavy filesystem activity makes these fail, and we can
+ * use compound pages.
+ */
+ gfp_flags = __GFP_WAIT | __GFP_IO | __GFP_COMP;
egrcnt = dd->ipath_rcvegrcnt;
/* TID number offset for this port */
@@ -721,10 +730,8 @@ static int ipath_create_user_egr(struct
* memory pressure (creating large files and then copying them over
* NFS while doing lots of MPI jobs), we hit some allocation
* failures, even though we can sleep... (2.6.10) Still get
- * failures at 64K. 32K is the lowest we can go without waiting
- * more memory again. It seems likely that the coalescing in
- * free_pages, etc. still has issues (as it has had previously
- * during 2.6.x development).
+ * failures at 64K. 32K is the lowest we can go without wasting
+ * additional memory.
*/
size = 0x8000;
alloced = ALIGN(egrsize * egrcnt, size);
@@ -745,12 +752,6 @@ static int ipath_create_user_egr(struct
goto bail_rcvegrbuf;
}
for (e = 0; e < pd->port_rcvegrbuf_chunks; e++) {
- /*
- * GFP_USER, but without GFP_FS, so buffer cache can be
- * coalesced (we hope); otherwise, even at order 4,
- * heavy filesystem activity makes these fail
- */
- gfp_t gfp_flags = __GFP_WAIT | __GFP_IO | __GFP_COMP;
pd->port_rcvegrbuf[e] = dma_alloc_coherent(
&dd->pcidev->dev, size, &pd->port_rcvegrbuf_phys[e],
@@ -1167,9 +1168,10 @@ static int ipath_mmap(struct file *fp, s
ureg = dd->ipath_uregbase + dd->ipath_palign * pd->port_port;
- ipath_cdbg(MM, "ushare: pgaddr %llx vm_start=%lx, vmlen %lx\n",
+ ipath_cdbg(MM, "pgaddr %llx vm_start=%lx len %lx port %u:%u\n",
(unsigned long long) pgaddr, vma->vm_start,
- vma->vm_end - vma->vm_start);
+ vma->vm_end - vma->vm_start, dd->ipath_unit,
+ pd->port_port);
if (pgaddr == ureg)
ret = mmap_ureg(vma, dd, ureg);
next prev parent reply other threads:[~2006-06-29 21:46 UTC|newest]
Thread overview: 65+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-06-29 21:40 [PATCH 0 of 39] ipath - bug fixes, performance enhancements, and portability improvements Bryan O'Sullivan
2006-06-29 21:40 ` [PATCH 1 of 39] IB/ipath - Name zero counter offsets so it's clear they aren't counters Bryan O'Sullivan
2006-06-29 21:40 ` [PATCH 2 of 39] IB/ipath - update copyrights and other strings to reflect new company name Bryan O'Sullivan
2006-06-29 21:40 ` [PATCH 3 of 39] IB/ipath - Share more common code between RC and UC protocols Bryan O'Sullivan
2006-06-29 21:40 ` [PATCH 4 of 39] IB/ipath - fix an indenting problem Bryan O'Sullivan
2006-06-29 21:40 ` [PATCH 5 of 39] IB/ipath - fix shared receive queues for RC Bryan O'Sullivan
2006-06-29 21:40 ` [PATCH 6 of 39] IB/ipath - Allow diags on any unit Bryan O'Sullivan
2006-06-29 21:40 ` [PATCH 7 of 39] IB/ipath - update some comments and fix typos Bryan O'Sullivan
2006-06-29 21:40 ` [PATCH 8 of 39] IB/ipath - remove some duplicate code Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 9 of 39] IB/ipath - don't allow resources to be created with illegal values Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 10 of 39] IB/ipath - fix some memory leaks on failure paths Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 11 of 39] IB/ipath - return an error for unknown multicast GID Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 12 of 39] IB/ipath - report correct device identification information in /sys Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 13 of 39] IB/ipath - enforce device resource limits Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 14 of 39] IB/ipath - removed unused field ipath_kregvirt from struct ipath_devdata Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 15 of 39] IB/ipath - print better debug info when handling 32/64-bit DMA mask problems Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 16 of 39] IB/ipath - enable freeze mode when shutting down device Bryan O'Sullivan
2006-06-29 21:41 ` Bryan O'Sullivan [this message]
2006-06-30 0:02 ` [PATCH 17 of 39] IB/ipath - use more appropriate gfp flags Andrew Morton
2006-06-29 21:41 ` [PATCH 18 of 39] IB/ipath - use vmalloc to allocate struct ipath_devdata Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 19 of 39] IB/ipath - memory management cleanups Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 20 of 39] IB/ipath - reduce overhead on receive interrupts Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 21 of 39] IB/ipath - fixed bug 9776 for real. The problem was that I was updating Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 22 of 39] IB/ipath - fix lost interrupts on HT-400 Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 23 of 39] IB/ipath - disallow send of invalid packet sizes over UD Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 24 of 39] IB/ipath - don't confuse the max message size with the MTU Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 25 of 39] IB/ipath - removed redundant statements Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 26 of 39] IB/ipath - check for valid LID and multicast LIDs Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 27 of 39] IB/ipath - fixes to performance get counters for IB compliance Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 28 of 39] IB/ipath - Fixes a bug where our delay for EEPROM no longer works due to compiler reordering Bryan O'Sullivan
2006-06-30 0:07 ` Andrew Morton
2006-06-30 21:08 ` Roland Dreier
2006-06-29 21:41 ` [PATCH 29 of 39] IB/ipath - RC receive interrupt performance changes Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 30 of 39] IB/ipath - purge sps_lid and sps_mlid arrays Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 31 of 39] IB/ipath - drop the "stats" sysfs attribute group Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 32 of 39] IB/ipath - support more models of InfiniPath hardware Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 33 of 39] IB/ipath - read/write correct sizes through diag interface Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 34 of 39] IB/ipath - fix a bug that results in addresses near 0 being written via DMA Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 35 of 39] IB/ipath - remove some #if 0 code related to lockable memory Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 36 of 39] IB/ipath - Ignore receive queue size if SRQ is specified Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 37 of 39] IB/ipath - namespace cleanup: replace ips with ipath Bryan O'Sullivan
2006-06-29 21:41 ` [PATCH 38 of 39] IB/ipath - More changes to support InfiniPath on PowerPC 970 systems Bryan O'Sullivan
2006-06-29 21:53 ` David Miller
2006-06-29 22:01 ` Bryan O'Sullivan
2006-06-29 22:04 ` David Miller
2006-07-03 22:25 ` Anton Blanchard
2006-07-04 4:55 ` Michael S. Tsirkin
2006-07-06 22:37 ` Benjamin Herrenschmidt
2006-07-06 23:34 ` Bryan O'Sullivan
2006-07-06 23:39 ` Benjamin Herrenschmidt
2006-06-29 21:41 ` [PATCH 39 of 39] IB/ipath - use streaming copy in RDMA interrupt handler to reduce packet loss Bryan O'Sullivan
2006-06-29 21:50 ` David Miller
2006-06-29 21:59 ` Bryan O'Sullivan
2006-06-29 22:03 ` David Miller
2006-06-29 23:34 ` Bryan O'Sullivan
2006-06-29 23:46 ` David Miller
2006-06-29 23:55 ` [openib-general] " Ralph Campbell
2006-06-30 0:28 ` Rick Jones
2006-06-30 0:32 ` David Miller
2006-06-30 0:44 ` Rick Jones
2006-06-30 0:47 ` David Miller
2006-06-30 16:31 ` [PATCH 0 of 39] ipath - bug fixes, performance enhancements,and portability improvements Michael S. Tsirkin
2006-06-30 17:00 ` Bryan O'Sullivan
2006-07-01 19:43 ` [PATCH 0 of 39] ipath - bug fixes, performanceenhancements,and " Michael S. Tsirkin
2006-07-01 22:35 ` Andrew Morton
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=9d943b828776136a2bb7.1151617268@eng-12.pathscale.com \
--to=bos@pathscale.com \
--cc=akpm@osdl.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mst@mellanox.co.il \
--cc=netdev@vger.kernel.org \
--cc=openib-general@openib.org \
--cc=rdreier@cisco.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
Powered by JetHome