From: Roland Dreier <roland@topspin.com>
To: akpm@osdl.org
Cc: linux-kernel@vger.kernel.org, openib-general@openib.org
Subject: [PATCH][23/26] IB/mthca: mem-free multicast table
Date: Thu, 3 Mar 2005 15:20:28 -0800 [thread overview]
Message-ID: <2005331520.kVkmRDQ3e4IStEy9@topspin.com> (raw)
In-Reply-To: <2005331520.ADYAIRdSQBiHhYiD@topspin.com>
Tie up one last loose end by mapping enough context memory to cover
the whole multicast table during initialization, and then enable
mem-free mode. mthca now supports enough of mem-free mode so that
IPoIB works with a mem-free HCA.
Signed-off-by: Roland Dreier <roland@topspin.com>
--- linux-export.orig/drivers/infiniband/hw/mthca/mthca_dev.h 2005-03-03 14:13:02.565340719 -0800
+++ linux-export/drivers/infiniband/hw/mthca/mthca_dev.h 2005-03-03 14:13:03.005245231 -0800
@@ -207,8 +207,9 @@
};
struct mthca_mcg_table {
- struct semaphore sem;
- struct mthca_alloc alloc;
+ struct semaphore sem;
+ struct mthca_alloc alloc;
+ struct mthca_icm_table *table;
};
struct mthca_dev {
--- linux-export.orig/drivers/infiniband/hw/mthca/mthca_main.c 2005-03-03 14:12:57.858362446 -0800
+++ linux-export/drivers/infiniband/hw/mthca/mthca_main.c 2005-03-03 14:13:03.005245231 -0800
@@ -412,8 +412,29 @@
goto err_unmap_eqp;
}
+ /*
+ * It's not strictly required, but for simplicity just map the
+ * whole multicast group table now. The table isn't very big
+ * and it's a lot easier than trying to track ref counts.
+ */
+ mdev->mcg_table.table = mthca_alloc_icm_table(mdev, init_hca->mc_base,
+ MTHCA_MGM_ENTRY_SIZE,
+ mdev->limits.num_mgms +
+ mdev->limits.num_amgms,
+ mdev->limits.num_mgms +
+ mdev->limits.num_amgms,
+ 0);
+ if (!mdev->mcg_table.table) {
+ mthca_err(mdev, "Failed to map MCG context memory, aborting.\n");
+ err = -ENOMEM;
+ goto err_unmap_cq;
+ }
+
return 0;
+err_unmap_cq:
+ mthca_free_icm_table(mdev, mdev->cq_table.table);
+
err_unmap_eqp:
mthca_free_icm_table(mdev, mdev->qp_table.eqp_table);
@@ -587,7 +608,7 @@
goto err_uar_free;
}
- err = mthca_init_pd_table(dev);
+ err = mthca_init_pd_table(dev);
if (err) {
mthca_err(dev, "Failed to initialize "
"protection domain table, aborting.\n");
@@ -635,13 +656,6 @@
mthca_dbg(dev, "NOP command IRQ test passed\n");
- if (dev->hca_type == ARBEL_NATIVE) {
- mthca_warn(dev, "Sorry, native MT25208 mode support is not complete, "
- "aborting.\n");
- err = -ENODEV;
- goto err_cmd_poll;
- }
-
err = mthca_init_cq_table(dev);
if (err) {
mthca_err(dev, "Failed to initialize "
@@ -704,7 +718,7 @@
err_uar_table_free:
mthca_cleanup_uar_table(dev);
- return err;
+ return err;
}
static int __devinit mthca_request_regions(struct pci_dev *pdev,
@@ -814,6 +828,7 @@
const struct pci_device_id *id)
{
static int mthca_version_printed = 0;
+ static int mthca_memfree_warned = 0;
int ddr_hidden = 0;
int err;
struct mthca_dev *mdev;
@@ -893,6 +908,10 @@
mdev->pdev = pdev;
mdev->hca_type = id->driver_data;
+ if (mdev->hca_type == ARBEL_NATIVE && !mthca_memfree_warned++)
+ mthca_warn(mdev, "Warning: native MT25208 mode support is incomplete. "
+ "Your HCA may not work properly.\n");
+
if (ddr_hidden)
mdev->mthca_flags |= MTHCA_FLAG_DDR_HIDDEN;
next prev parent reply other threads:[~2005-03-03 23:30 UTC|newest]
Thread overview: 47+ messages / expand[flat|nested] mbox.gz Atom feed top
2005-03-03 23:20 [PATCH][0/26] InfiniBand merge Roland Dreier
2005-03-03 23:20 ` [PATCH][1/26] IB: fix ib_find_cached_gid() port numbering Roland Dreier
2005-03-03 23:20 ` [PATCH][2/26] IB/mthca: CQ minor tweaks Roland Dreier
2005-03-03 23:20 ` [PATCH][3/26] IB/mthca: improve CQ locking part 1 Roland Dreier
2005-03-03 23:20 ` [PATCH][4/26] IB/mthca: improve CQ locking part 2 Roland Dreier
2005-03-03 23:20 ` [PATCH][5/26] IB/mthca: CQ cleanups Roland Dreier
2005-03-03 23:20 ` [PATCH][6/26] IB: remove unsignaled receives Roland Dreier
2005-03-03 23:20 ` [PATCH][7/26] IB/mthca: map registers for mem-free mode Roland Dreier
2005-03-03 23:20 ` [PATCH][8/26] IB/mthca: add UAR allocation Roland Dreier
2005-03-03 23:20 ` [PATCH][9/26] IB/mthca: dynamic context memory mapping for mem-free mode Roland Dreier
2005-03-03 23:20 ` [PATCH][10/26] IB/mthca: mem-free memory region support Roland Dreier
2005-03-03 23:20 ` [PATCH][11/26] IB/mthca: mem-free EQ initialization Roland Dreier
2005-03-03 23:20 ` [PATCH][12/26] IB/mthca: mem-free interrupt handling Roland Dreier
2005-03-03 23:20 ` [PATCH][13/26] IB/mthca: tweak firmware command debug messages Roland Dreier
2005-03-03 23:20 ` [PATCH][14/26] IB/mthca: tweak MAP_ICM_page firmware command Roland Dreier
2005-03-03 23:20 ` [PATCH][15/26] IB/mthca: mem-free doorbell record allocation Roland Dreier
2005-03-03 23:20 ` [PATCH][16/26] IB/mthca: mem-free doorbell record writing Roland Dreier
2005-03-03 23:20 ` [PATCH][17/26] IB/mthca: refactor CQ buffer allocate/free Roland Dreier
2005-03-03 23:20 ` [PATCH][18/26] IB/mthca: mem-free CQ initialization Roland Dreier
2005-03-03 23:20 ` [PATCH][19/26] IB/mthca: mem-free CQ operations Roland Dreier
2005-03-03 23:20 ` [PATCH][20/26] IB/mthca: mem-free QP initialization Roland Dreier
2005-03-03 23:20 ` [PATCH][21/26] IB/mthca: mem-free address vectors Roland Dreier
2005-03-03 23:20 ` [PATCH][22/26] IB/mthca: mem-free work request posting Roland Dreier
2005-03-03 23:20 ` Roland Dreier [this message]
2005-03-03 23:20 ` [PATCH][24/26] IB/mthca: QP locking optimization Roland Dreier
2005-03-03 23:20 ` [PATCH][25/26] IB/mthca: implement query of device caps Roland Dreier
2005-03-03 23:20 ` [PATCH][26/26] IB: MAD cancel callbacks from thread Roland Dreier
2005-03-04 0:07 ` Jeff Garzik
2005-03-04 0:30 ` Roland Dreier
2005-03-04 0:34 ` [openib-general] Re: [PATCH][26/26] IB: MAD cancel callbacks fromthread Sean Hefty
2005-03-04 0:43 ` Roland Dreier
2005-03-04 1:01 ` Andrew Morton
2005-03-04 1:07 ` Andrew Morton
2005-03-04 1:22 ` Andrew Morton
2006-03-13 15:43 ` [PATCH 5/6 v2] IB: IP address based RDMA connection manager Roland Dreier
2006-03-13 17:11 ` Sean Hefty
2006-03-13 17:26 ` Roland Dreier
2005-03-04 0:04 ` [PATCH][16/26] IB/mthca: mem-free doorbell record writing Jeff Garzik
2005-03-04 0:33 ` Roland Dreier
2005-03-04 0:41 ` Jeff Garzik
2005-03-04 0:50 ` Roland Dreier
2005-03-04 0:35 ` [PATCH][3/26] IB/mthca: improve CQ locking part 1 Jeff Garzik
2005-03-04 0:40 ` Roland Dreier
2005-03-04 0:58 ` Greg KH
2005-03-04 1:02 ` Roland Dreier
2005-03-04 16:33 ` Greg KH
2005-03-04 16:43 ` Roland Dreier
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=2005331520.kVkmRDQ3e4IStEy9@topspin.com \
--to=roland@topspin.com \
--cc=akpm@osdl.org \
--cc=linux-kernel@vger.kernel.org \
--cc=openib-general@openib.org \
/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
all inboxes | Powered by JetHome®