From: Ben Hutchings <bhutchings@solarflare.com>
To: David Miller <davem@davemloft.net>
Cc: netdev@vger.kernel.org, linux-net-drivers@solarflare.com
Subject: [PATCH 07/27] sfc: Move RX data FIFO thresholds out of struct efx_nic_type
Date: Fri, 23 Oct 2009 19:30:17 +0100 [thread overview]
Message-ID: <1256322617.2785.17.camel@achroite> (raw)
In-Reply-To: <1256322441.2785.3.camel@achroite>
Since there are now separate blocks of code to set the thresholds for
each NIC type, it is no longer useful to include them in the NIC type
description.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
---
drivers/net/sfc/falcon.c | 23 ++++++++++++-----------
drivers/net/sfc/net_driver.h | 4 ----
2 files changed, 12 insertions(+), 15 deletions(-)
diff --git a/drivers/net/sfc/falcon.c b/drivers/net/sfc/falcon.c
index 4cb98d4..c23e8e2 100644
--- a/drivers/net/sfc/falcon.c
+++ b/drivers/net/sfc/falcon.c
@@ -2922,16 +2922,17 @@ static void falcon_init_rx_cfg(struct efx_nic *efx)
const unsigned ctrl_xon_thr = 20;
const unsigned ctrl_xoff_thr = 25;
/* RX data FIFO thresholds (256-byte units; size varies) */
- unsigned data_xon_thr =
- ((rx_xon_thresh_bytes >= 0) ?
- rx_xon_thresh_bytes : efx->type->rx_xon_thresh) >> 8;
- unsigned data_xoff_thr =
- ((rx_xoff_thresh_bytes >= 0) ?
- rx_xoff_thresh_bytes : efx->type->rx_xoff_thresh) >> 8;
+ int data_xon_thr = rx_xon_thresh_bytes >> 8;
+ int data_xoff_thr = rx_xoff_thresh_bytes >> 8;
efx_oword_t reg;
falcon_read(efx, ®, RX_CFG_REG_KER);
if (falcon_rev(efx) <= FALCON_REV_A1) {
+ /* Data FIFO size is 5.5K */
+ if (data_xon_thr < 0)
+ data_xon_thr = 512 >> 8;
+ if (data_xoff_thr < 0)
+ data_xoff_thr = 2048 >> 8;
EFX_SET_OWORD_FIELD(reg, RX_DESC_PUSH_EN_A1, 0);
EFX_SET_OWORD_FIELD(reg, RX_USR_BUF_SIZE_A1, huge_buf_size);
EFX_SET_OWORD_FIELD(reg, RX_XON_MAC_TH_A1, data_xon_thr);
@@ -2939,7 +2940,11 @@ static void falcon_init_rx_cfg(struct efx_nic *efx)
EFX_SET_OWORD_FIELD(reg, RX_XON_TX_TH_A1, ctrl_xon_thr);
EFX_SET_OWORD_FIELD(reg, RX_XOFF_TX_TH_A1, ctrl_xoff_thr);
} else {
- /* Register fields moved */
+ /* Data FIFO size is 80K; register fields moved */
+ if (data_xon_thr < 0)
+ data_xon_thr = 27648 >> 8; /* ~3*max MTU */
+ if (data_xoff_thr < 0)
+ data_xoff_thr = 54272 >> 8; /* ~80Kb - 3*max MTU */
EFX_SET_OWORD_FIELD(reg, RX_DESC_PUSH_EN_B0, 0);
EFX_SET_OWORD_FIELD(reg, RX_USR_BUF_SIZE_B0, huge_buf_size);
EFX_SET_OWORD_FIELD(reg, RX_XON_MAC_TH_B0, data_xon_thr);
@@ -3130,8 +3135,6 @@ struct efx_nic_type falcon_a_nic_type = {
.max_dma_mask = FALCON_DMA_MASK,
.tx_dma_mask = FALCON_TX_DMA_MASK,
.bug5391_mask = 0xf,
- .rx_xoff_thresh = 2048,
- .rx_xon_thresh = 512,
.rx_buffer_padding = 0x24,
.max_interrupt_mode = EFX_INT_MODE_MSI,
.phys_addr_channels = 4,
@@ -3154,8 +3157,6 @@ struct efx_nic_type falcon_b_nic_type = {
.max_dma_mask = FALCON_DMA_MASK,
.tx_dma_mask = FALCON_TX_DMA_MASK,
.bug5391_mask = 0,
- .rx_xoff_thresh = 54272, /* ~80Kb - 3*max MTU */
- .rx_xon_thresh = 27648, /* ~3*max MTU */
.rx_buffer_padding = 0,
.max_interrupt_mode = EFX_INT_MODE_MSIX,
.phys_addr_channels = 32, /* Hardware limit is 64, but the legacy
diff --git a/drivers/net/sfc/net_driver.h b/drivers/net/sfc/net_driver.h
index 298566d..479a6fe 100644
--- a/drivers/net/sfc/net_driver.h
+++ b/drivers/net/sfc/net_driver.h
@@ -875,8 +875,6 @@ static inline const char *efx_dev_name(struct efx_nic *efx)
* @max_dma_mask: Maximum possible DMA mask
* @tx_dma_mask: TX DMA mask
* @bug5391_mask: Address mask for bug 5391 workaround
- * @rx_xoff_thresh: RX FIFO XOFF watermark (bytes)
- * @rx_xon_thresh: RX FIFO XON watermark (bytes)
* @rx_buffer_padding: Padding added to each RX buffer
* @max_interrupt_mode: Highest capability interrupt mode supported
* from &enum efx_init_mode.
@@ -899,8 +897,6 @@ struct efx_nic_type {
unsigned int tx_dma_mask;
unsigned bug5391_mask;
- int rx_xoff_thresh;
- int rx_xon_thresh;
unsigned int rx_buffer_padding;
unsigned int max_interrupt_mode;
unsigned int phys_addr_channels;
--
Ben Hutchings, Senior Software Engineer, Solarflare Communications
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
next prev parent reply other threads:[~2009-10-23 18:30 UTC|newest]
Thread overview: 29+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-10-23 18:27 [PATCH 00/27] sfc: Cleanup and preparation for new hardware Ben Hutchings
2009-10-23 18:28 ` [PATCH 01/27] sfc: Remove redundant header gmii.h Ben Hutchings
2009-10-23 18:28 ` [PATCH 02/27] sfc: Remove redundant hardware initialisation Ben Hutchings
2009-10-23 18:29 ` [PATCH 03/27] sfc: Rename Falcon-specific board code and types Ben Hutchings
2009-10-23 18:29 ` [PATCH 04/27] sfc: Merge sfe4001.c into falcon_boards.c Ben Hutchings
2009-10-23 18:29 ` [PATCH 05/27] sfc: Remove boards.h, moving last remaining declaration to falcon.h Ben Hutchings
2009-10-23 18:30 ` [PATCH 06/27] sfc: Remove versioned bitfield macros Ben Hutchings
2009-10-23 18:30 ` Ben Hutchings [this message]
2009-10-23 18:30 ` [PATCH 08/27] sfc: Update hardware definitions for Siena Ben Hutchings
2009-10-23 18:30 ` [PATCH 09/27] sfc: Rename register I/O header and functions used by both Falcon and Siena Ben Hutchings
2009-10-23 18:30 ` [PATCH 10/27] sfc: Eliminate indirect lookups of queue size constants Ben Hutchings
2009-10-23 18:31 ` [PATCH 11/27] sfc: Define DMA address mask explicitly in terms of descriptor field width Ben Hutchings
2009-10-23 18:31 ` [PATCH 12/27] sfc: Move all TX DMA length limiting into tx.c Ben Hutchings
2009-10-23 18:31 ` [PATCH 13/27] sfc: Merge struct efx_blinker into struct efx_board Ben Hutchings
2009-10-23 18:31 ` [PATCH 14/27] sfc: Change order of device removal to reverse of probe order Ben Hutchings
2009-10-23 18:31 ` [PATCH 15/27] sfc: Remove declarations of nonexistent functions Ben Hutchings
2009-10-23 18:31 ` [PATCH 16/27] sfc: Move efx_xmit_done() declaration into correct stanza Ben Hutchings
2009-10-23 18:32 ` [PATCH 17/27] sfc: Move shared members of struct falcon_nic_data into struct efx_nic Ben Hutchings
2009-10-23 18:32 ` [PATCH 18/27] sfc: Maintain interrupt moderation values in ticks, not microseconds Ben Hutchings
2009-10-23 18:32 ` [PATCH 19/27] sfc: Removed kernel-doc for nonexistent member of efx_phy_operations Ben Hutchings
2009-10-23 18:32 ` [PATCH 20/27] sfc: Remove pointless abstraction of memory BAR number Ben Hutchings
2009-10-23 18:32 ` [PATCH 21/27] sfc: Merge falcon_probe_phy() into falcon_probe_port() Ben Hutchings
2009-10-23 18:32 ` [PATCH 22/27] sfc: Remove incorrect assertion from efx_pci_remove_main() Ben Hutchings
2009-10-23 18:33 ` [PATCH 23/27] sfc: Remove unnecessary tests of efx->membase Ben Hutchings
2009-10-23 18:33 ` [PATCH 24/27] sfc: Move MTD probe after netdev registration and name allocation Ben Hutchings
2009-10-23 18:33 ` [PATCH 25/27] sfc: Merge efx_fc_resolve() into efx_mdio_get_pause() Ben Hutchings
2009-10-23 18:33 ` [PATCH 26/27] sfc: Remove unused code for non-autoneg speed/duplex switching Ben Hutchings
2009-10-23 18:33 ` [PATCH 27/27] sfc: Rename 'xfp' file and functions to reflect reality Ben Hutchings
2009-10-24 11:29 ` [PATCH 00/27] sfc: Cleanup and preparation for new hardware David Miller
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=1256322617.2785.17.camel@achroite \
--to=bhutchings@solarflare.com \
--cc=davem@davemloft.net \
--cc=linux-net-drivers@solarflare.com \
--cc=netdev@vger.kernel.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.