UbiquitousOS
/
OpenCloudOS-Kernel
mirror of https://gitee.com/OpenCloudOS/OpenCloudOS-Kernel.git
11
0
Fork 0
Code Issues Projects Releases Wiki Activity
OpenCloudOS-Kernel/drivers/net/sgiseeq.c

857 lines
22 KiB
C
Raw Normal View History

Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/*
* sgiseeq.c: Seeq8003 ethernet driver for SGI machines.
*
* Copyright (C) 1996 David S. Miller (dm@engr.sgi.com)
*/
[PATCH] Sparse: Cleanup sgiseeq sparse warnings. o Make sgiseeq_dump_rings static. o Delete unused sgiseeq_my_reset. o Move DEBUG define to beginning where it's easier to spot and will be seen by <linux/kernel.h> as well. o Use NULL for pointer initialization. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-03-08 19:49:31 +08:00
#undef DEBUG
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/errno.h>
#include <linux/init.h>
#include <linux/types.h>
#include <linux/interrupt.h>
#include <linux/string.h>
#include <linux/delay.h>
#include <linux/netdevice.h>
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
#include <linux/platform_device.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
#include <linux/etherdevice.h>
#include <linux/skbuff.h>
#include <asm/sgi/hpc3.h>
#include <asm/sgi/ip22.h>
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
#include <asm/sgi/seeq.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
#include "sgiseeq.h"
static char *sgiseeqstr = "SGI Seeq8003";
/*
* If you want speed, you do something silly, it always has worked for me. So,
* with that in mind, I've decided to make this driver look completely like a
* stupid Lance from a driver architecture perspective. Only difference is that
* here our "ring buffer" looks and acts like a real Lance one does but is
* layed out like how the HPC DMA and the Seeq want it to. You'd be surprised
* how a stupid idea like this can pay off in performance, not to mention
* making this driver 2,000 times easier to write. ;-)
*/
/* Tune these if we tend to run out often etc. */
#define SEEQ_RX_BUFFERS 16
#define SEEQ_TX_BUFFERS 16
#define PKT_BUF_SZ 1584
#define NEXT_RX(i) (((i) + 1) & (SEEQ_RX_BUFFERS - 1))
#define NEXT_TX(i) (((i) + 1) & (SEEQ_TX_BUFFERS - 1))
#define PREV_RX(i) (((i) - 1) & (SEEQ_RX_BUFFERS - 1))
#define PREV_TX(i) (((i) - 1) & (SEEQ_TX_BUFFERS - 1))
#define TX_BUFFS_AVAIL(sp) ((sp->tx_old <= sp->tx_new) ? \
sp->tx_old + (SEEQ_TX_BUFFERS - 1) - sp->tx_new : \
sp->tx_old - sp->tx_new - 1)
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
#define VIRT_TO_DMA(sp, v) ((sp)->srings_dma + \
(dma_addr_t)((unsigned long)(v) - \
(unsigned long)((sp)->rx_desc)))
/* Copy frames shorter than rx_copybreak, otherwise pass on up in
* a full sized sk_buff. Value of 100 stolen from tulip.c (!alpha).
*/
static int rx_copybreak = 100;
#define PAD_SIZE (128 - sizeof(struct hpc_dma_desc) - sizeof(void *))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
struct sgiseeq_rx_desc {
volatile struct hpc_dma_desc rdma;
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
u8 padding[PAD_SIZE];
struct sk_buff *skb;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
};
struct sgiseeq_tx_desc {
volatile struct hpc_dma_desc tdma;
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
u8 padding[PAD_SIZE];
struct sk_buff *skb;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
};
/*
* Warning: This structure is layed out in a certain way because HPC dma
* descriptors must be 8-byte aligned. So don't touch this without
* some care.
*/
struct sgiseeq_init_block { /* Note the name ;-) */
struct sgiseeq_rx_desc rxvector[SEEQ_RX_BUFFERS];
struct sgiseeq_tx_desc txvector[SEEQ_TX_BUFFERS];
};
struct sgiseeq_private {
struct sgiseeq_init_block *srings;
sgiseeq: replace use of dma_cache_wback_inv The sgiseeq driver is one of the few remaining users of the ancient cache banging DMA API. Replaced with the modern days DMA API. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-09-04 21:41:01 +08:00
dma_addr_t srings_dma;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* Ptrs to the descriptors in uncached space. */
struct sgiseeq_rx_desc *rx_desc;
struct sgiseeq_tx_desc *tx_desc;
char *name;
struct hpc3_ethregs *hregs;
struct sgiseeq_regs *sregs;
/* Ring entry counters. */
unsigned int rx_new, tx_new;
unsigned int rx_old, tx_old;
int is_edlc;
unsigned char control;
unsigned char mode;
spinlock_t tx_lock;
};
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
static inline void dma_sync_desc_cpu(struct net_device *dev, void *addr)
{
dma_cache_sync(dev->dev.parent, addr, sizeof(struct sgiseeq_rx_desc),
DMA_FROM_DEVICE);
}
static inline void dma_sync_desc_dev(struct net_device *dev, void *addr)
{
dma_cache_sync(dev->dev.parent, addr, sizeof(struct sgiseeq_rx_desc),
DMA_TO_DEVICE);
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
static inline void hpc3_eth_reset(struct hpc3_ethregs *hregs)
{
[PATCH] sgiseeq: Configure PIO and DMA timing requests. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> drivers/net/sgiseeq.c | 28 ++++++++++++++-------------- include/asm-mips/sgi/hpc3.h | 40 ++++++++++++++++++++-------------------- 2 files changed, 34 insertions(+), 34 deletions(-) Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-10 21:50:56 +08:00
hregs->reset = HPC3_ERST_CRESET | HPC3_ERST_CLRIRQ;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
udelay(20);
[PATCH] sgiseeq: Configure PIO and DMA timing requests. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> drivers/net/sgiseeq.c | 28 ++++++++++++++-------------- include/asm-mips/sgi/hpc3.h | 40 ++++++++++++++++++++-------------------- 2 files changed, 34 insertions(+), 34 deletions(-) Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-10 21:50:56 +08:00
hregs->reset = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
static inline void reset_hpc3_and_seeq(struct hpc3_ethregs *hregs,
struct sgiseeq_regs *sregs)
{
hregs->rx_ctrl = hregs->tx_ctrl = 0;
hpc3_eth_reset(hregs);
}
#define RSTAT_GO_BITS (SEEQ_RCMD_IGOOD | SEEQ_RCMD_IEOF | SEEQ_RCMD_ISHORT | \
SEEQ_RCMD_IDRIB | SEEQ_RCMD_ICRC)
static inline void seeq_go(struct sgiseeq_private *sp,
struct hpc3_ethregs *hregs,
struct sgiseeq_regs *sregs)
{
sregs->rstat = sp->mode | RSTAT_GO_BITS;
hregs->rx_ctrl = HPC3_ERXCTRL_ACTIVE;
}
static inline void __sgiseeq_set_mac_address(struct net_device *dev)
{
struct sgiseeq_private *sp = netdev_priv(dev);
struct sgiseeq_regs *sregs = sp->sregs;
int i;
sregs->tstat = SEEQ_TCMD_RB0;
for (i = 0; i < 6; i++)
sregs->rw.eth_addr[i] = dev->dev_addr[i];
}
static int sgiseeq_set_mac_address(struct net_device *dev, void *addr)
{
struct sgiseeq_private *sp = netdev_priv(dev);
struct sockaddr *sa = addr;
memcpy(dev->dev_addr, sa->sa_data, dev->addr_len);
spin_lock_irq(&sp->tx_lock);
__sgiseeq_set_mac_address(dev);
spin_unlock_irq(&sp->tx_lock);
return 0;
}
#define TCNTINFO_INIT (HPCDMA_EOX | HPCDMA_ETXD)
#define RCNTCFG_INIT (HPCDMA_OWN | HPCDMA_EORP | HPCDMA_XIE)
#define RCNTINFO_INIT (RCNTCFG_INIT | (PKT_BUF_SZ & HPCDMA_BCNT))
static int seeq_init_ring(struct net_device *dev)
{
struct sgiseeq_private *sp = netdev_priv(dev);
int i;
netif_stop_queue(dev);
sp->rx_new = sp->tx_new = 0;
sp->rx_old = sp->tx_old = 0;
__sgiseeq_set_mac_address(dev);
/* Setup tx ring. */
for(i = 0; i < SEEQ_TX_BUFFERS; i++) {
sp->tx_desc[i].tdma.cntinfo = TCNTINFO_INIT;
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, &sp->tx_desc[i]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
/* And now the rx ring. */
for (i = 0; i < SEEQ_RX_BUFFERS; i++) {
SGISEEQ: fix oops when doing ifconfig down; ifconfig up When doing init_ring checking whether a new skb needs to be allocated was wrong. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2008-01-13 07:08:47 +08:00
if (!sp->rx_desc[i].skb) {
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
dma_addr_t dma_addr;
struct sk_buff *skb = netdev_alloc_skb(dev, PKT_BUF_SZ);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
if (skb == NULL)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return -ENOMEM;
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
skb_reserve(skb, 2);
dma_addr = dma_map_single(dev->dev.parent,
skb->data - 2,
PKT_BUF_SZ, DMA_FROM_DEVICE);
sp->rx_desc[i].skb = skb;
sp->rx_desc[i].rdma.pbuf = dma_addr;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
sp->rx_desc[i].rdma.cntinfo = RCNTINFO_INIT;
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, &sp->rx_desc[i]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
sp->rx_desc[i - 1].rdma.cntinfo |= HPCDMA_EOR;
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, &sp->rx_desc[i - 1]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return 0;
}
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
static void seeq_purge_ring(struct net_device *dev)
{
struct sgiseeq_private *sp = netdev_priv(dev);
int i;
/* clear tx ring. */
for (i = 0; i < SEEQ_TX_BUFFERS; i++) {
if (sp->tx_desc[i].skb) {
dev_kfree_skb(sp->tx_desc[i].skb);
sp->tx_desc[i].skb = NULL;
}
}
/* And now the rx ring. */
for (i = 0; i < SEEQ_RX_BUFFERS; i++) {
if (sp->rx_desc[i].skb) {
dev_kfree_skb(sp->rx_desc[i].skb);
sp->rx_desc[i].skb = NULL;
}
}
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
#ifdef DEBUG
static struct sgiseeq_private *gpriv;
static struct net_device *gdev;
[PATCH] Sparse: Cleanup sgiseeq sparse warnings. o Make sgiseeq_dump_rings static. o Delete unused sgiseeq_my_reset. o Move DEBUG define to beginning where it's easier to spot and will be seen by <linux/kernel.h> as well. o Use NULL for pointer initialization. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-03-08 19:49:31 +08:00
static void sgiseeq_dump_rings(void)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
static int once;
struct sgiseeq_rx_desc *r = gpriv->rx_desc;
struct sgiseeq_tx_desc *t = gpriv->tx_desc;
struct hpc3_ethregs *hregs = gpriv->hregs;
int i;
if (once)
return;
once++;
printk("RING DUMP:\n");
for (i = 0; i < SEEQ_RX_BUFFERS; i++) {
printk("RX [%d]: @(%p) [%08x,%08x,%08x] ",
i, (&r[i]), r[i].rdma.pbuf, r[i].rdma.cntinfo,
r[i].rdma.pnext);
i += 1;
printk("-- [%d]: @(%p) [%08x,%08x,%08x]\n",
i, (&r[i]), r[i].rdma.pbuf, r[i].rdma.cntinfo,
r[i].rdma.pnext);
}
for (i = 0; i < SEEQ_TX_BUFFERS; i++) {
printk("TX [%d]: @(%p) [%08x,%08x,%08x] ",
i, (&t[i]), t[i].tdma.pbuf, t[i].tdma.cntinfo,
t[i].tdma.pnext);
i += 1;
printk("-- [%d]: @(%p) [%08x,%08x,%08x]\n",
i, (&t[i]), t[i].tdma.pbuf, t[i].tdma.cntinfo,
t[i].tdma.pnext);
}
printk("INFO: [rx_new = %d rx_old=%d] [tx_new = %d tx_old = %d]\n",
gpriv->rx_new, gpriv->rx_old, gpriv->tx_new, gpriv->tx_old);
printk("RREGS: rx_cbptr[%08x] rx_ndptr[%08x] rx_ctrl[%08x]\n",
hregs->rx_cbptr, hregs->rx_ndptr, hregs->rx_ctrl);
printk("TREGS: tx_cbptr[%08x] tx_ndptr[%08x] tx_ctrl[%08x]\n",
hregs->tx_cbptr, hregs->tx_ndptr, hregs->tx_ctrl);
}
#endif
#define TSTAT_INIT_SEEQ (SEEQ_TCMD_IPT|SEEQ_TCMD_I16|SEEQ_TCMD_IC|SEEQ_TCMD_IUF)
#define TSTAT_INIT_EDLC ((TSTAT_INIT_SEEQ) | SEEQ_TCMD_RB2)
static int init_seeq(struct net_device *dev, struct sgiseeq_private *sp,
struct sgiseeq_regs *sregs)
{
struct hpc3_ethregs *hregs = sp->hregs;
int err;
reset_hpc3_and_seeq(hregs, sregs);
err = seeq_init_ring(dev);
if (err)
return err;
/* Setup to field the proper interrupt types. */
if (sp->is_edlc) {
sregs->tstat = TSTAT_INIT_EDLC;
sregs->rw.wregs.control = sp->control;
sregs->rw.wregs.frame_gap = 0;
} else {
sregs->tstat = TSTAT_INIT_SEEQ;
}
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
hregs->rx_ndptr = VIRT_TO_DMA(sp, sp->rx_desc);
hregs->tx_ndptr = VIRT_TO_DMA(sp, sp->tx_desc);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
seeq_go(sp, hregs, sregs);
return 0;
}
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
static void record_rx_errors(struct net_device *dev, unsigned char status)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
if (status & SEEQ_RSTAT_OVERF ||
status & SEEQ_RSTAT_SFRAME)
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
dev->stats.rx_over_errors++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (status & SEEQ_RSTAT_CERROR)
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
dev->stats.rx_crc_errors++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (status & SEEQ_RSTAT_DERROR)
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
dev->stats.rx_frame_errors++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (status & SEEQ_RSTAT_REOF)
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
dev->stats.rx_errors++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
static inline void rx_maybe_restart(struct sgiseeq_private *sp,
struct hpc3_ethregs *hregs,
struct sgiseeq_regs *sregs)
{
if (!(hregs->rx_ctrl & HPC3_ERXCTRL_ACTIVE)) {
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
hregs->rx_ndptr = VIRT_TO_DMA(sp, sp->rx_desc + sp->rx_new);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
seeq_go(sp, hregs, sregs);
}
}
static inline void sgiseeq_rx(struct net_device *dev, struct sgiseeq_private *sp,
struct hpc3_ethregs *hregs,
struct sgiseeq_regs *sregs)
{
struct sgiseeq_rx_desc *rd;
[PATCH] Sparse: Cleanup sgiseeq sparse warnings. o Make sgiseeq_dump_rings static. o Delete unused sgiseeq_my_reset. o Move DEBUG define to beginning where it's easier to spot and will be seen by <linux/kernel.h> as well. o Use NULL for pointer initialization. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2006-03-08 19:49:31 +08:00
struct sk_buff *skb = NULL;
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
struct sk_buff *newskb;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
unsigned char pkt_status;
int len = 0;
unsigned int orig_end = PREV_RX(sp->rx_new);
/* Service every received packet. */
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
rd = &sp->rx_desc[sp->rx_new];
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_cpu(dev, rd);
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
while (!(rd->rdma.cntinfo & HPCDMA_OWN)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
len = PKT_BUF_SZ - (rd->rdma.cntinfo & HPCDMA_BCNT) - 3;
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
dma_unmap_single(dev->dev.parent, rd->rdma.pbuf,
PKT_BUF_SZ, DMA_FROM_DEVICE);
pkt_status = rd->skb->data[len];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (pkt_status & SEEQ_RSTAT_FIG) {
/* Packet is OK. */
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
/* We don't want to receive our own packets */
if (memcmp(rd->skb->data + 6, dev->dev_addr, ETH_ALEN)) {
if (len > rx_copybreak) {
skb = rd->skb;
newskb = netdev_alloc_skb(dev, PKT_BUF_SZ);
if (!newskb) {
newskb = skb;
skb = NULL;
goto memory_squeeze;
}
skb_reserve(newskb, 2);
} else {
net: Use netdev_alloc_skb_ip_align() Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-13 13:34:20 +08:00
skb = netdev_alloc_skb_ip_align(dev, len);
if (skb)
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
skb_copy_to_linear_data(skb, rd->skb->data, len);
net: Use netdev_alloc_skb_ip_align() Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-13 13:34:20 +08:00
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
newskb = rd->skb;
}
memory_squeeze:
if (skb) {
skb_put(skb, len);
skb->protocol = eth_type_trans(skb, dev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
netif_rx(skb);
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
dev->stats.rx_packets++;
dev->stats.rx_bytes += len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
} else {
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
printk(KERN_NOTICE "%s: Memory squeeze, deferring packet.\n",
dev->name);
dev->stats.rx_dropped++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
} else {
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
/* Silently drop my own packets */
newskb = rd->skb;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
} else {
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
record_rx_errors(dev, pkt_status);
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
newskb = rd->skb;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
rd->skb = newskb;
rd->rdma.pbuf = dma_map_single(dev->dev.parent,
newskb->data - 2,
PKT_BUF_SZ, DMA_FROM_DEVICE);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* Return the entry to the ring pool. */
rd->rdma.cntinfo = RCNTINFO_INIT;
sp->rx_new = NEXT_RX(sp->rx_new);
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, rd);
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
rd = &sp->rx_desc[sp->rx_new];
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_cpu(dev, rd);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_cpu(dev, &sp->rx_desc[orig_end]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
sp->rx_desc[orig_end].rdma.cntinfo &= ~(HPCDMA_EOR);
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, &sp->rx_desc[orig_end]);
dma_sync_desc_cpu(dev, &sp->rx_desc[PREV_RX(sp->rx_new)]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
sp->rx_desc[PREV_RX(sp->rx_new)].rdma.cntinfo |= HPCDMA_EOR;
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, &sp->rx_desc[PREV_RX(sp->rx_new)]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
rx_maybe_restart(sp, hregs, sregs);
}
static inline void tx_maybe_reset_collisions(struct sgiseeq_private *sp,
struct sgiseeq_regs *sregs)
{
if (sp->is_edlc) {
sregs->rw.wregs.control = sp->control & ~(SEEQ_CTRL_XCNT);
sregs->rw.wregs.control = sp->control;
}
}
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
static inline void kick_tx(struct net_device *dev,
struct sgiseeq_private *sp,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
struct hpc3_ethregs *hregs)
{
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
struct sgiseeq_tx_desc *td;
int i = sp->tx_old;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* If the HPC aint doin nothin, and there are more packets
* with ETXD cleared and XIU set we must make very certain
* that we restart the HPC else we risk locking up the
* adapter. The following code is only safe iff the HPCDMA
* is not active!
*/
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
td = &sp->tx_desc[i];
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_cpu(dev, td);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
while ((td->tdma.cntinfo & (HPCDMA_XIU | HPCDMA_ETXD)) ==
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
(HPCDMA_XIU | HPCDMA_ETXD)) {
i = NEXT_TX(i);
td = &sp->tx_desc[i];
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_cpu(dev, td);
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (td->tdma.cntinfo & HPCDMA_XIU) {
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
hregs->tx_ndptr = VIRT_TO_DMA(sp, td);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
hregs->tx_ctrl = HPC3_ETXCTRL_ACTIVE;
}
}
static inline void sgiseeq_tx(struct net_device *dev, struct sgiseeq_private *sp,
struct hpc3_ethregs *hregs,
struct sgiseeq_regs *sregs)
{
struct sgiseeq_tx_desc *td;
unsigned long status = hregs->tx_ctrl;
int j;
tx_maybe_reset_collisions(sp, sregs);
if (!(status & (HPC3_ETXCTRL_ACTIVE | SEEQ_TSTAT_PTRANS))) {
/* Oops, HPC detected some sort of error. */
if (status & SEEQ_TSTAT_R16)
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
dev->stats.tx_aborted_errors++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (status & SEEQ_TSTAT_UFLOW)
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
dev->stats.tx_fifo_errors++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (status & SEEQ_TSTAT_LCLS)
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
dev->stats.collisions++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
/* Ack 'em... */
for (j = sp->tx_old; j != sp->tx_new; j = NEXT_TX(j)) {
td = &sp->tx_desc[j];
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_cpu(dev, td);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (!(td->tdma.cntinfo & (HPCDMA_XIU)))
break;
if (!(td->tdma.cntinfo & (HPCDMA_ETXD))) {
if (!(status & HPC3_ETXCTRL_ACTIVE)) {
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
hregs->tx_ndptr = VIRT_TO_DMA(sp, td);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
hregs->tx_ctrl = HPC3_ETXCTRL_ACTIVE;
}
break;
}
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
dev->stats.tx_packets++;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
sp->tx_old = NEXT_TX(sp->tx_old);
td->tdma.cntinfo &= ~(HPCDMA_XIU | HPCDMA_XIE);
td->tdma.cntinfo |= HPCDMA_EOX;
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
if (td->skb) {
dev_kfree_skb_any(td->skb);
td->skb = NULL;
}
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, td);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
}
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 21:55:46 +08:00
static irqreturn_t sgiseeq_interrupt(int irq, void *dev_id)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
struct net_device *dev = (struct net_device *) dev_id;
struct sgiseeq_private *sp = netdev_priv(dev);
struct hpc3_ethregs *hregs = sp->hregs;
struct sgiseeq_regs *sregs = sp->sregs;
spin_lock(&sp->tx_lock);
/* Ack the IRQ and set software state. */
[PATCH] sgiseeq: Configure PIO and DMA timing requests. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> drivers/net/sgiseeq.c | 28 ++++++++++++++-------------- include/asm-mips/sgi/hpc3.h | 40 ++++++++++++++++++++-------------------- 2 files changed, 34 insertions(+), 34 deletions(-) Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-10 21:50:56 +08:00
hregs->reset = HPC3_ERST_CLRIRQ;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* Always check for received packets. */
sgiseeq_rx(dev, sp, hregs, sregs);
/* Only check for tx acks if we have something queued. */
if (sp->tx_old != sp->tx_new)
sgiseeq_tx(dev, sp, hregs, sregs);
if ((TX_BUFFS_AVAIL(sp) > 0) && netif_queue_stopped(dev)) {
netif_wake_queue(dev);
}
spin_unlock(&sp->tx_lock);
return IRQ_HANDLED;
}
static int sgiseeq_open(struct net_device *dev)
{
struct sgiseeq_private *sp = netdev_priv(dev);
struct sgiseeq_regs *sregs = sp->sregs;
unsigned int irq = dev->irq;
int err;
if (request_irq(irq, sgiseeq_interrupt, 0, sgiseeqstr, dev)) {
printk(KERN_ERR "Seeq8003: Can't get irq %d\n", dev->irq);
err = -EAGAIN;
}
err = init_seeq(dev, sp, sregs);
if (err)
goto out_free_irq;
netif_start_queue(dev);
return 0;
out_free_irq:
free_irq(irq, dev);
return err;
}
static int sgiseeq_close(struct net_device *dev)
{
struct sgiseeq_private *sp = netdev_priv(dev);
struct sgiseeq_regs *sregs = sp->sregs;
[PATCH] sgiseeq: Fix resource handling. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> drivers/net/sgiseeq.c | 9 ++++----- 1 files changed, 4 insertions(+), 5 deletions(-) Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-10 21:50:51 +08:00
unsigned int irq = dev->irq;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
netif_stop_queue(dev);
/* Shutdown the Seeq. */
reset_hpc3_and_seeq(sp->hregs, sregs);
[PATCH] sgiseeq: Fix resource handling. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> drivers/net/sgiseeq.c | 9 ++++----- 1 files changed, 4 insertions(+), 5 deletions(-) Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-10 21:50:51 +08:00
free_irq(irq, dev);
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
seeq_purge_ring(dev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return 0;
}
static inline int sgiseeq_reset(struct net_device *dev)
{
struct sgiseeq_private *sp = netdev_priv(dev);
struct sgiseeq_regs *sregs = sp->sregs;
int err;
err = init_seeq(dev, sp, sregs);
if (err)
return err;
dev->trans_start = jiffies;
netif_wake_queue(dev);
return 0;
}
static int sgiseeq_start_xmit(struct sk_buff *skb, struct net_device *dev)
{
struct sgiseeq_private *sp = netdev_priv(dev);
struct hpc3_ethregs *hregs = sp->hregs;
unsigned long flags;
struct sgiseeq_tx_desc *td;
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
int len, entry;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
spin_lock_irqsave(&sp->tx_lock, flags);
/* Setup... */
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
len = skb->len;
if (len < ETH_ZLEN) {
if (skb_padto(skb, ETH_ZLEN))
net: use NETDEV_TX_OK instead of 0 in ndo_start_xmit() functions This patch is the result of an automatic spatch transformation to convert all ndo_start_xmit() return values of 0 to NETDEV_TX_OK. Some occurences are missed by the automatic conversion, those will be handled in a seperate patch. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-23 14:03:08 +08:00
return NETDEV_TX_OK;
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
len = ETH_ZLEN;
}
[NET] drivers/net: statistics cleanup #1 -- save memory and shrink code We now have struct net_device_stats embedded in struct net_device, and the default ->get_stats() hook does the obvious thing for us. Run through drivers/net/* and remove the driver-local storage of statistics, and driver-local ->get_stats() hook where applicable. This was just the low-hanging fruit in drivers/net; plenty more drivers remain to be updated. [ Resolved conflicts with napi_struct changes and fix sunqe build regression... -DaveM ] Signed-off-by: Jeff Garzik <jeff@garzik.org> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-10-04 08:41:50 +08:00
dev->stats.tx_bytes += len;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
entry = sp->tx_new;
td = &sp->tx_desc[entry];
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_cpu(dev, td);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* Create entry. There are so many races with adding a new
* descriptor to the chain:
* 1) Assume that the HPC is off processing a DMA chain while
* we are changing all of the following.
* 2) Do no allow the HPC to look at a new descriptor until
* we have completely set up it's state. This means, do
* not clear HPCDMA_EOX in the current last descritptor
* until the one we are adding looks consistent and could
* be processes right now.
* 3) The tx interrupt code must notice when we've added a new
* entry and the HPC got to the end of the chain before we
* added this new entry and restarted it.
*/
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
td->skb = skb;
td->tdma.pbuf = dma_map_single(dev->dev.parent, skb->data,
len, DMA_TO_DEVICE);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
td->tdma.cntinfo = (len & HPCDMA_BCNT) |
HPCDMA_XIU | HPCDMA_EOXP | HPCDMA_XIE | HPCDMA_EOX;
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, td);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (sp->tx_old != sp->tx_new) {
struct sgiseeq_tx_desc *backend;
backend = &sp->tx_desc[PREV_TX(sp->tx_new)];
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_cpu(dev, backend);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
backend->tdma.cntinfo &= ~HPCDMA_EOX;
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, backend);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
sp->tx_new = NEXT_TX(sp->tx_new); /* Advance. */
/* Maybe kick the HPC back into motion. */
if (!(hregs->tx_ctrl & HPC3_ETXCTRL_ACTIVE))
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
kick_tx(dev, sp, hregs);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
dev->trans_start = jiffies;
if (!TX_BUFFS_AVAIL(sp))
netif_stop_queue(dev);
spin_unlock_irqrestore(&sp->tx_lock, flags);
net: use NETDEV_TX_OK instead of 0 in ndo_start_xmit() functions This patch is the result of an automatic spatch transformation to convert all ndo_start_xmit() return values of 0 to NETDEV_TX_OK. Some occurences are missed by the automatic conversion, those will be handled in a seperate patch. Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-06-23 14:03:08 +08:00
return NETDEV_TX_OK;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
static void timeout(struct net_device *dev)
{
printk(KERN_NOTICE "%s: transmit timed out, resetting\n", dev->name);
sgiseeq_reset(dev);
dev->trans_start = jiffies;
netif_wake_queue(dev);
}
static void sgiseeq_set_multicast(struct net_device *dev)
{
netdevice: safe convert to netdev_priv() #part-3 We have some reasons to kill netdev->priv: 1. netdev->priv is equal to netdev_priv(). 2. netdev_priv() wraps the calculation of netdev->priv's offset, obviously netdev_priv() is more flexible than netdev->priv. But we cann't kill netdev->priv, because so many drivers reference to it directly. This patch is a safe convert for netdev->priv to netdev_priv(netdev). Since all of the netdev->priv is only for read. But it is too big to be sent in one mail. I split it to 4 parts and make every part smaller than 100,000 bytes, which is max size allowed by vger. Signed-off-by: Wang Chen <wangchen@cn.fujitsu.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-11-13 15:38:36 +08:00
struct sgiseeq_private *sp = netdev_priv(dev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
unsigned char oldmode = sp->mode;
if(dev->flags & IFF_PROMISC)
sp->mode = SEEQ_RCMD_RANY;
else if ((dev->flags & IFF_ALLMULTI) || dev->mc_count)
sp->mode = SEEQ_RCMD_RBMCAST;
else
sp->mode = SEEQ_RCMD_RBCAST;
/* XXX I know this sucks, but is there a better way to reprogram
* XXX the receiver? At least, this shouldn't happen too often.
*/
if (oldmode != sp->mode)
sgiseeq_reset(dev);
}
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
static inline void setup_tx_ring(struct net_device *dev,
struct sgiseeq_tx_desc *buf,
int nbufs)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
struct sgiseeq_private *sp = netdev_priv(dev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
int i = 0;
while (i < (nbufs - 1)) {
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
buf[i].tdma.pnext = VIRT_TO_DMA(sp, buf + i + 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
buf[i].tdma.pbuf = 0;
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, &buf[i]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
i++;
}
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
buf[i].tdma.pnext = VIRT_TO_DMA(sp, buf);
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, &buf[i]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
static inline void setup_rx_ring(struct net_device *dev,
struct sgiseeq_rx_desc *buf,
int nbufs)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
struct sgiseeq_private *sp = netdev_priv(dev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
int i = 0;
while (i < (nbufs - 1)) {
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
buf[i].rdma.pnext = VIRT_TO_DMA(sp, buf + i + 1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
buf[i].rdma.pbuf = 0;
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, &buf[i]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
i++;
}
buf[i].rdma.pbuf = 0;
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
buf[i].rdma.pnext = VIRT_TO_DMA(sp, buf);
SGISEEQ: use cached memory access to make driver work on IP28 - Use inline functions for dma_sync_* instead of macros - added Kconfig change to make selection for similair SGI boxes easier Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-12-19 20:42:36 +08:00
dma_sync_desc_dev(dev, &buf[i]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
sgiseeq: convert to net_device_ops Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-15 20:52:58 +08:00
static const struct net_device_ops sgiseeq_netdev_ops = {
.ndo_open = sgiseeq_open,
.ndo_stop = sgiseeq_close,
.ndo_start_xmit = sgiseeq_start_xmit,
.ndo_tx_timeout = timeout,
.ndo_set_multicast_list = sgiseeq_set_multicast,
.ndo_set_mac_address = sgiseeq_set_mac_address,
.ndo_change_mtu = eth_change_mtu,
.ndo_validate_addr = eth_validate_addr,
};
net: move sgiseeq's probe function to .devinit.text A pointer to sgiseeq_probe is passed to the core via platform_driver_register and so the function must not disappear when the .init sections are discarded. Otherwise (if also having HOTPLUG=y) unbinding and binding a device to the driver via sysfs will result in an oops as does a device being registered late. An alternative to this patch is using platform_driver_probe instead of platform_driver_register plus removing the pointer to the probe function from the struct platform_driver. Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Cc: Wang Chen <wangchen@cn.fujitsu.com> Cc: David S. Miller <davem@davemloft.net> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: Jeff Garzik <jgarzik@redhat.com> Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Acked-by: Ralf Baechle <ralf@linux-mips.org> Cc: Greg Kroah-Hartman <gregkh@suse.de> Cc: netdev@vger.kernel.org Signed-off-by: David S. Miller <davem@davemloft.net>
2009-07-21 19:25:26 +08:00
static int __devinit sgiseeq_probe(struct platform_device *pdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
struct sgiseeq_platform_data *pd = pdev->dev.platform_data;
struct hpc3_regs *hpcregs = pd->hpc;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
struct sgiseeq_init_block *sr;
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
unsigned int irq = pd->irq;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
struct sgiseeq_private *sp;
struct net_device *dev;
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
int err;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
dev = alloc_etherdev(sizeof (struct sgiseeq_private));
if (!dev) {
printk(KERN_ERR "Sgiseeq: Etherdev alloc failed, aborting.\n");
err = -ENOMEM;
goto err_out;
}
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
platform_set_drvdata(pdev, dev);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
sp = netdev_priv(dev);
/* Make private data page aligned */
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
sr = dma_alloc_noncoherent(&pdev->dev, sizeof(*sp->srings),
sgiseeq: replace use of dma_cache_wback_inv The sgiseeq driver is one of the few remaining users of the ancient cache banging DMA API. Replaced with the modern days DMA API. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-09-04 21:41:01 +08:00
&sp->srings_dma, GFP_KERNEL);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (!sr) {
printk(KERN_ERR "Sgiseeq: Page alloc failed, aborting.\n");
err = -ENOMEM;
goto err_out_free_dev;
}
sp->srings = sr;
sgiseeq: replace use of dma_cache_wback_inv The sgiseeq driver is one of the few remaining users of the ancient cache banging DMA API. Replaced with the modern days DMA API. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-09-04 21:41:01 +08:00
sp->rx_desc = sp->srings->rxvector;
sp->tx_desc = sp->srings->txvector;
/* A couple calculations now, saves many cycles later. */
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
setup_rx_ring(dev, sp->rx_desc, SEEQ_RX_BUFFERS);
setup_tx_ring(dev, sp->tx_desc, SEEQ_TX_BUFFERS);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
memcpy(dev->dev_addr, pd->mac, ETH_ALEN);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
#ifdef DEBUG
gpriv = sp;
gdev = dev;
#endif
[PATCH] sgiseeq: Configure PIO and DMA timing requests. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> drivers/net/sgiseeq.c | 28 ++++++++++++++-------------- include/asm-mips/sgi/hpc3.h | 40 ++++++++++++++++++++-------------------- 2 files changed, 34 insertions(+), 34 deletions(-) Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-10 21:50:56 +08:00
sp->sregs = (struct sgiseeq_regs *) &hpcregs->eth_ext[0];
sp->hregs = &hpcregs->ethregs;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
sp->name = sgiseeqstr;
sp->mode = SEEQ_RCMD_RBCAST;
Add support for Seeq 8003 on Challenge S Mezz board. Thanks to Jö Fahlke for donating hardware. Signed-off-by: Ladislav Michl <ladis@linux-mips.org> Forward porting of Ladis' 2.4 patch. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-28 09:18:35 +08:00
/* Setup PIO and DMA transfer timing */
sp->hregs->pconfig = 0x161;
sp->hregs->dconfig = HPC3_EDCFG_FIRQ | HPC3_EDCFG_FEOP |
HPC3_EDCFG_FRXDC | HPC3_EDCFG_PTO | 0x026;
[PATCH] sgiseeq: Configure PIO and DMA timing requests. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> drivers/net/sgiseeq.c | 28 ++++++++++++++-------------- include/asm-mips/sgi/hpc3.h | 40 ++++++++++++++++++++-------------------- 2 files changed, 34 insertions(+), 34 deletions(-) Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-10 21:50:56 +08:00
/* Setup PIO and DMA transfer timing */
sp->hregs->pconfig = 0x161;
sp->hregs->dconfig = HPC3_EDCFG_FIRQ | HPC3_EDCFG_FEOP |
HPC3_EDCFG_FRXDC | HPC3_EDCFG_PTO | 0x026;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* Reset the chip. */
hpc3_eth_reset(sp->hregs);
sp->is_edlc = !(sp->sregs->rw.rregs.collision_tx[0] & 0xff);
if (sp->is_edlc)
sp->control = SEEQ_CTRL_XCNT | SEEQ_CTRL_ACCNT |
SEEQ_CTRL_SFLAG | SEEQ_CTRL_ESHORT |
SEEQ_CTRL_ENCARR;
sgiseeq: convert to net_device_ops Signed-off-by: Alexander Beregalov <a.beregalov@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2009-04-15 20:52:58 +08:00
dev->netdev_ops = &sgiseeq_netdev_ops;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
dev->watchdog_timeo = (200 * HZ) / 1000;
dev->irq = irq;
if (register_netdev(dev)) {
printk(KERN_ERR "Sgiseeq: Cannot register net device, "
"aborting.\n");
err = -ENODEV;
goto err_out_free_page;
}
net: convert print_mac to %pM This converts pretty much everything to print_mac. There were a few things that had conflicts which I have just dropped for now, no harm done. I've built an allyesconfig with this and looked at the files that weren't built very carefully, but it's a huge patch. Signed-off-by: Johannes Berg <johannes@sipsolutions.net> Signed-off-by: David S. Miller <davem@davemloft.net>
2008-10-28 06:59:26 +08:00
printk(KERN_INFO "%s: %s %pM\n", dev->name, sgiseeqstr, dev->dev_addr);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return 0;
err_out_free_page:
[PATCH] sgiseeq: Fix resource handling. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> drivers/net/sgiseeq.c | 9 ++++----- 1 files changed, 4 insertions(+), 5 deletions(-) Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-10 21:50:51 +08:00
free_page((unsigned long) sp->srings);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
err_out_free_dev:
kfree(dev);
err_out:
return err;
}
sgiseeq: Fix return type of sgiseeq_remove The driver remove method needs to return an int not void. This was just never noticed because usually this driver is not being built as a module. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-22 23:03:52 +08:00
static int __exit sgiseeq_remove(struct platform_device *pdev)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
struct net_device *dev = platform_get_drvdata(pdev);
struct sgiseeq_private *sp = netdev_priv(dev);
Add support for Seeq 8003 on Challenge S Mezz board. Thanks to Jö Fahlke for donating hardware. Signed-off-by: Ladislav Michl <ladis@linux-mips.org> Forward porting of Ladis' 2.4 patch. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-02-28 09:18:35 +08:00
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
unregister_netdev(dev);
SGISEEQ: use cached memory access to make driver work on IP28 SGI IP28 machines would need special treatment (enable adding addtional wait states) when accessing memory uncached. To avoid this pain I changed the driver to use only cached access to memory. Signed-off-by: Thomas Bogendoerfer <tsbogend@alpha.franken.de> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-11-24 20:29:19 +08:00
dma_free_noncoherent(&pdev->dev, sizeof(*sp->srings), sp->srings,
sp->srings_dma);
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
free_netdev(dev);
platform_set_drvdata(pdev, NULL);
sgiseeq: Fix return type of sgiseeq_remove The driver remove method needs to return an int not void. This was just never noticed because usually this driver is not being built as a module. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Jeff Garzik <jeff@garzik.org>
2007-08-22 23:03:52 +08:00
return 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
static struct platform_driver sgiseeq_driver = {
.probe = sgiseeq_probe,
don't use __devexit_p to wrap sgiseeq_remove The function sgiseeq_remove is defined using __exit, so don't use __devexit_p but __exit_p to wrap it. Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de> Cc: David S. Miller <davem@davemloft.net> Cc: Wang Chen <wangchen@cn.fujitsu.com> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Patrick McHardy <kaber@trash.net> Cc: netdev@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: David S. Miller <davem@davemloft.net>
2009-10-01 06:28:26 +08:00
.remove = __exit_p(sgiseeq_remove),
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
.driver = {
net drivers: fix platform driver hotplug/coldplug Since 43cc71eed1250755986da4c0f9898f9a635cb3bf, the platform modalias is prefixed with "platform:". Add MODULE_ALIAS() to the hotpluggable network platform drivers, to re-enable auto loading. NOTE: didn't change drivers/net/fs_enet/fs_enet-main.c "old binding" support. That looks problematic in the first place (it even uses the ancient "struct device_driver" binding scheme for platform_bus!) and I suspect it will vanish soonish when arch/powerpc rules the world. Also, drivers/net/ne.c would have needed more thought to sort out. [akpm@linux-foundation.org: fix sgiseeq.c] [dbrownell@users.sourceforge.net: more drivers, registration fixes] Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Scott Wood <scottwood@freescale.com> Cc: Vitaly Bordug <vitb@kernel.crashing.org> Cc: Dale Farnsworth <dale@farnsworth.org> Cc: Ben Dooks <ben-linux@fluff.org> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Andrew Victor <andrew@sanpeople.com> Cc: Bryan Wu <bryan.wu@analog.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-19 04:50:44 +08:00
.name = "sgiseeq",
.owner = THIS_MODULE,
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
}
};
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
static int __init sgiseeq_module_init(void)
{
if (platform_driver_register(&sgiseeq_driver)) {
printk(KERN_ERR "Driver registration failed\n");
return -ENODEV;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
return 0;
}
static void __exit sgiseeq_module_exit(void)
{
platform_driver_unregister(&sgiseeq_driver);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
Convert SGI IP22 and specific drivers to platform_device. Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2007-05-11 22:48:58 +08:00
module_init(sgiseeq_module_init);
module_exit(sgiseeq_module_exit);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
[PATCH] sgiseeq: Configure PIO and DMA timing requests. Signed-off-by: Ralf Baechle <ralf@linux-mips.org> drivers/net/sgiseeq.c | 28 ++++++++++++++-------------- include/asm-mips/sgi/hpc3.h | 40 ++++++++++++++++++++-------------------- 2 files changed, 34 insertions(+), 34 deletions(-) Signed-off-by: Jeff Garzik <jgarzik@pobox.com>
2005-10-10 21:50:56 +08:00
MODULE_DESCRIPTION("SGI Seeq 8003 driver");
MODULE_AUTHOR("Linux/MIPS Mailing List <linux-mips@linux-mips.org>");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
MODULE_LICENSE("GPL");
net drivers: fix platform driver hotplug/coldplug Since 43cc71eed1250755986da4c0f9898f9a635cb3bf, the platform modalias is prefixed with "platform:". Add MODULE_ALIAS() to the hotpluggable network platform drivers, to re-enable auto loading. NOTE: didn't change drivers/net/fs_enet/fs_enet-main.c "old binding" support. That looks problematic in the first place (it even uses the ancient "struct device_driver" binding scheme for platform_bus!) and I suspect it will vanish soonish when arch/powerpc rules the world. Also, drivers/net/ne.c would have needed more thought to sort out. [akpm@linux-foundation.org: fix sgiseeq.c] [dbrownell@users.sourceforge.net: more drivers, registration fixes] Signed-off-by: Kay Sievers <kay.sievers@vrfy.org> Signed-off-by: David Brownell <dbrownell@users.sourceforge.net> Cc: Scott Wood <scottwood@freescale.com> Cc: Vitaly Bordug <vitb@kernel.crashing.org> Cc: Dale Farnsworth <dale@farnsworth.org> Cc: Ben Dooks <ben-linux@fluff.org> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Andrew Victor <andrew@sanpeople.com> Cc: Bryan Wu <bryan.wu@analog.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
2008-04-19 04:50:44 +08:00
MODULE_ALIAS("platform:sgiseeq");