mirrors/linux-stable
1
0
Fork 0
mirror of https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git synced 2024-09-18 00:24:39 +00:00
Code Issues Releases Wiki Activity

Merge commit 'v2.6.37-rc4' into imx-for-2.6.38

Browse source 
Done to resolve merge conflict:

Conflicts:
	arch/arm/mach-mx25/devices-imx25.h

Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This commit is contained in:
Sascha Hauer 2010-12-06 09:36:17 +01:00
commit 0e44e05958
351 changed files with 10762 additions and 3123 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

4
Documentation/DocBook/sh.tmpl
View file

@ -79,10 +79,6 @@
</sect2>
</sect1>
</chapter>
<chapter id="clk">
<title>Clock Framework Extensions</title>
!Iinclude/linux/sh_clk.h
</chapter>
<chapter id="mach">
<title>Machine Specific Interfaces</title>
<sect1 id="dreamcast">

8
Documentation/edac.txt
View file

@ -196,7 +196,7 @@ csrow3.
The representation of the above is reflected in the directory tree
in EDAC's sysfs interface. Starting in directory
/sys/devices/system/edac/mc each memory controller will be represented
by its own 'mcX' directory, where 'X" is the index of the MC.
by its own 'mcX' directory, where 'X' is the index of the MC.
..../edac/mc/
@ -207,7 +207,7 @@ by its own 'mcX' directory, where 'X" is the index of the MC.
....
Under each 'mcX' directory each 'csrowX' is again represented by a
'csrowX', where 'X" is the csrow index:
'csrowX', where 'X' is the csrow index:
.../mc/mc0/
@ -232,7 +232,7 @@ EDAC control and attribute files.
In 'mcX' directories are EDAC control and attribute files for
this 'X" instance of the memory controllers:
this 'X' instance of the memory controllers:
Counter reset control file:
@ -343,7 +343,7 @@ Sdram memory scrubbing rate:
'csrowX' DIRECTORIES
In the 'csrowX' directories are EDAC control and attribute files for
this 'X" instance of csrow:
this 'X' instance of csrow:
Total Uncorrectable Errors count attribute file:

32
Documentation/fb/00-INDEX
View file

@ -4,33 +4,41 @@ please mail me.
Geert Uytterhoeven <geert@linux-m68k.org>
00-INDEX
- this file
- this file.
arkfb.txt
- info on the fbdev driver for ARK Logic chips.
aty128fb.txt
- info on the ATI Rage128 frame buffer driver.
cirrusfb.txt
- info on the driver for Cirrus Logic chipsets.
cmap_xfbdev.txt
- an introduction to fbdev's cmap structures.
deferred_io.txt
- an introduction to deferred IO.
efifb.txt
- info on the EFI platform driver for Intel based Apple computers.
ep93xx-fb.txt
- info on the driver for EP93xx LCD controller.
fbcon.txt
- intro to and usage guide for the framebuffer console (fbcon).
framebuffer.txt
- introduction to frame buffer devices.
imacfb.txt
- info on the generic EFI platform driver for Intel based Macs.
gxfb.txt
- info on the framebuffer driver for AMD Geode GX2 based processors.
intel810.txt
- documentation for the Intel 810/815 framebuffer driver.
intelfb.txt
- docs for Intel 830M/845G/852GM/855GM/865G/915G/945G fb driver.
internals.txt
- quick overview of frame buffer device internals.
lxfb.txt
- info on the framebuffer driver for AMD Geode LX based processors.
matroxfb.txt
- info on the Matrox framebuffer driver for Alpha, Intel and PPC.
metronomefb.txt
- info on the driver for the Metronome display controller.
modedb.txt
- info on the video mode database.
matroxfb.txt
- info on the Matrox frame buffer driver.
pvr2fb.txt
- info on the PowerVR 2 frame buffer driver.
pxafb.txt
@ -39,13 +47,23 @@ s3fb.txt
- info on the fbdev driver for S3 Trio/Virge chips.
sa1100fb.txt
- information about the driver for the SA-1100 LCD controller.
sh7760fb.txt
- info on the SH7760/SH7763 integrated LCDC Framebuffer driver.
sisfb.txt
- info on the framebuffer device driver for various SiS chips.
sstfb.txt
- info on the frame buffer driver for 3dfx' Voodoo Graphics boards.
tgafb.txt
- info on the TGA (DECChip 21030) frame buffer driver
- info on the TGA (DECChip 21030) frame buffer driver.
tridentfb.txt
info on the framebuffer driver for some Trident chip based cards.
uvesafb.txt
- info on the userspace VESA (VBE2+ compliant) frame buffer device.
vesafb.txt
- info on the VESA frame buffer device
- info on the VESA frame buffer device.
viafb.modes
- list of modes for VIA Integration Graphic Chip.
viafb.txt
- info on the VIA Integration Graphic Chip console framebuffer driver.
vt8623fb.txt
- info on the fb driver for the graphics core in VIA VT8623 chipsets.

5
Documentation/kernel-parameters.txt
View file

@ -2385,6 +2385,11 @@ and is between 256 and 4096 characters. It is defined in the file
improve throughput, but will also increase the
amount of memory reserved for use by the client.
swapaccount[=0|1]
[KNL] Enable accounting of swap in memory resource
controller if no parameter or 1 is given or disable
it if 0 is given (See Documentation/cgroups/memory.txt)
swiotlb= [IA-64] Number of I/O TLB slabs
switches= [HW,M68k]

1
Documentation/networking/ip-sysctl.txt
View file

@ -144,6 +144,7 @@ tcp_adv_win_scale - INTEGER
Count buffering overhead as bytes/2^tcp_adv_win_scale
(if tcp_adv_win_scale > 0) or bytes-bytes/2^(-tcp_adv_win_scale),
if it is <= 0.
Possible values are [-31, 31], inclusive.
Default: 2
tcp_allowed_congestion_control - STRING

32
Documentation/sh/clk.txt
View file

@ -1,32 +0,0 @@
Clock framework on SuperH architecture
The framework on SH extends existing API by the function clk_set_rate_ex,
which prototype is as follows:
clk_set_rate_ex (struct clk *clk, unsigned long rate, int algo_id)
The algo_id parameter is used to specify algorithm used to recalculate clocks,
adjanced to clock, specified as first argument. It is assumed that algo_id==0
means no changes to adjanced clock
Internally, the clk_set_rate_ex forwards request to clk->ops->set_rate method,
if it is present in ops structure. The method should set the clock rate and adjust
all needed clocks according to the passed algo_id.
Exact values for algo_id are machine-dependent. For the sh7722, the following
values are defined:
NO_CHANGE = 0,
IUS_N1_N1, /* I:U = N:1, U:Sh = N:1 */
IUS_322, /* I:U:Sh = 3:2:2 */
IUS_522, /* I:U:Sh = 5:2:2 */
IUS_N11, /* I:U:Sh = N:1:1 */
SB_N1, /* Sh:B = N:1 */
SB3_N1, /* Sh:B3 = N:1 */
SB3_32, /* Sh:B3 = 3:2 */
SB3_43, /* Sh:B3 = 4:3 */
SB3_54, /* Sh:B3 = 5:4 */
BP_N1, /* B:P = N:1 */
IP_N1 /* I:P = N:1 */
Each of these constants means relation between clocks that can be set via the FRQCR
register

8
MAINTAINERS
View file

@ -1359,7 +1359,7 @@ F: include/net/bluetooth/
BONDING DRIVER
M: Jay Vosburgh <fubar@us.ibm.com>
L: bonding-devel@lists.sourceforge.net
L: netdev@vger.kernel.org
W: http://sourceforge.net/projects/bonding/
S: Supported
F: drivers/net/bonding/
@ -2444,10 +2444,12 @@ F: drivers/net/wan/sdla.c
FRAMEBUFFER LAYER
L: linux-fbdev@vger.kernel.org
W: http://linux-fbdev.sourceforge.net/
Q: http://patchwork.kernel.org/project/linux-fbdev/list/
T: git git://git.kernel.org/pub/scm/linux/kernel/git/lethal/fbdev-2.6.git
S: Orphan
F: Documentation/fb/
F: drivers/video/fb*
F: drivers/video/
F: include/video/
F: include/linux/fb.h
FREESCALE DMA DRIVER
@ -5837,6 +5839,8 @@ M: Chris Metcalf <cmetcalf@tilera.com>
W: http://www.tilera.com/scm/
S: Supported
F: arch/tile/
F: drivers/char/hvc_tile.c
F: drivers/net/tile/
TLAN NETWORK DRIVER
M: Samuel Chessman <chessman@tux.org>

2
Makefile
View file

@ -1,7 +1,7 @@
VERSION = 2
PATCHLEVEL = 6
SUBLEVEL = 37
EXTRAVERSION = -rc3
EXTRAVERSION = -rc4
NAME = Flesh-Eating Bats with Fangs
# *DOCUMENTATION*

2
arch/arm/boot/compressed/head.S
View file

@ -1084,6 +1084,6 @@ memdump: mov r12, r0
reloc_end:
.align
.section ".stack", "w"
.section ".stack", "aw", %nobits
user_stack: .space 4096
user_stack_end:

2
arch/arm/boot/compressed/vmlinux.lds.in
View file

@ -57,7 +57,7 @@ SECTIONS
.bss : { *(.bss) }
_end = .;
.stack (NOLOAD) : { *(.stack) }
.stack : { *(.stack) }
.stab 0 : { *(.stab) }
.stabstr 0 : { *(.stabstr) }

2
arch/arm/include/asm/assembler.h
View file

@ -238,7 +238,7 @@
@ Slightly optimised to avoid incrementing the pointer twice
usraccoff \instr, \reg, \ptr, \inc, 0, \cond, \abort
.if \rept == 2
usraccoff \instr, \reg, \ptr, \inc, 4, \cond, \abort
usraccoff \instr, \reg, \ptr, \inc, \inc, \cond, \abort
.endif
add\cond \ptr, #\rept * \inc

4
arch/arm/include/asm/mmu.h
View file

@ -13,6 +13,10 @@ typedef struct {
#ifdef CONFIG_CPU_HAS_ASID
#define ASID(mm) ((mm)->context.id & 255)
/* init_mm.context.id_lock should be initialized. */
#define INIT_MM_CONTEXT(name) \
.context.id_lock = __SPIN_LOCK_UNLOCKED(name.context.id_lock),
#else
#define ASID(mm) (0)
#endif

3
arch/arm/include/asm/pgtable.h
View file

@ -374,6 +374,9 @@ static inline pte_t *pmd_page_vaddr(pmd_t pmd)
#define pmd_page(pmd) pfn_to_page(__phys_to_pfn(pmd_val(pmd)))
/* we don't need complex calculations here as the pmd is folded into the pgd */
#define pmd_addr_end(addr,end) (end)
/*
* Conversion functions: convert a page and protection to a page entry,
* and a page entry and page directory to the page they refer to.

6
arch/arm/lib/findbit.S
View file

@ -174,8 +174,8 @@ ENDPROC(_find_next_bit_be)
*/
.L_found:
#if __LINUX_ARM_ARCH__ >= 5
rsb r1, r3, #0
and r3, r3, r1
rsb r0, r3, #0
and r3, r3, r0
clz r3, r3
rsb r3, r3, #31
add r0, r2, r3
@ -190,5 +190,7 @@ ENDPROC(_find_next_bit_be)
addeq r2, r2, #1
mov r0, r2
#endif
cmp r1, r0 @ Clamp to maxbit
movlo r0, r1
mov pc, lr

2
arch/arm/mach-aaec2000/include/mach/vmalloc.h
View file

@ -11,6 +11,6 @@
#ifndef __ASM_ARCH_VMALLOC_H
#define __ASM_ARCH_VMALLOC_H
#define VMALLOC_END 0xd0000000
#define VMALLOC_END 0xd0000000UL
#endif /* __ASM_ARCH_VMALLOC_H */

2
arch/arm/mach-bcmring/include/mach/vmalloc.h
View file

@ -22,4 +22,4 @@
* 0xe0000000 to 0xefffffff. This gives us 256 MB of vm space and handles
* larger physical memory designs better.
*/
#define VMALLOC_END 0xf0000000
#define VMALLOC_END 0xf0000000UL

2
arch/arm/mach-clps711x/include/mach/vmalloc.h
View file

@ -17,4 +17,4 @@
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
#define VMALLOC_END 0xd0000000
#define VMALLOC_END 0xd0000000UL

6
arch/arm/mach-davinci/dm355.c
View file

@ -359,8 +359,8 @@ static struct clk_lookup dm355_clks[] = {
CLK(NULL, "uart1", &uart1_clk),
CLK(NULL, "uart2", &uart2_clk),
CLK("i2c_davinci.1", NULL, &i2c_clk),
CLK("davinci-asp.0", NULL, &asp0_clk),
CLK("davinci-asp.1", NULL, &asp1_clk),
CLK("davinci-mcbsp.0", NULL, &asp0_clk),
CLK("davinci-mcbsp.1", NULL, &asp1_clk),
CLK("davinci_mmc.0", NULL, &mmcsd0_clk),
CLK("davinci_mmc.1", NULL, &mmcsd1_clk),
CLK("spi_davinci.0", NULL, &spi0_clk),
@ -664,7 +664,7 @@ static struct resource dm355_asp1_resources[] = {
};
static struct platform_device dm355_asp1_device = {
.name = "davinci-asp",
.name = "davinci-mcbsp",
.id = 1,
.num_resources = ARRAY_SIZE(dm355_asp1_resources),
.resource = dm355_asp1_resources,

6
arch/arm/mach-davinci/dm365.c
View file

@ -459,7 +459,7 @@ static struct clk_lookup dm365_clks[] = {
CLK(NULL, "usb", &usb_clk),
CLK("davinci_emac.1", NULL, &emac_clk),
CLK("davinci_voicecodec", NULL, &voicecodec_clk),
CLK("davinci-asp.0", NULL, &asp0_clk),
CLK("davinci-mcbsp", NULL, &asp0_clk),
CLK(NULL, "rto", &rto_clk),
CLK(NULL, "mjcp", &mjcp_clk),
CLK(NULL, NULL, NULL),
@ -922,8 +922,8 @@ static struct resource dm365_asp_resources[] = {
};
static struct platform_device dm365_asp_device = {
.name = "davinci-asp",
.id = 0,
.name = "davinci-mcbsp",
.id = -1,
.num_resources = ARRAY_SIZE(dm365_asp_resources),
.resource = dm365_asp_resources,
};

4
arch/arm/mach-davinci/dm644x.c
View file

@ -302,7 +302,7 @@ static struct clk_lookup dm644x_clks[] = {
CLK("davinci_emac.1", NULL, &emac_clk),
CLK("i2c_davinci.1", NULL, &i2c_clk),
CLK("palm_bk3710", NULL, &ide_clk),
CLK("davinci-asp", NULL, &asp_clk),
CLK("davinci-mcbsp", NULL, &asp_clk),
CLK("davinci_mmc.0", NULL, &mmcsd_clk),
CLK(NULL, "spi", &spi_clk),
CLK(NULL, "gpio", &gpio_clk),
@ -580,7 +580,7 @@ static struct resource dm644x_asp_resources[] = {
};
static struct platform_device dm644x_asp_device = {
.name = "davinci-asp",
.name = "davinci-mcbsp",
.id = -1,
.num_resources = ARRAY_SIZE(dm644x_asp_resources),
.resource = dm644x_asp_resources,

2
arch/arm/mach-ebsa110/include/mach/vmalloc.h
View file

@ -7,4 +7,4 @@
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation.
*/
#define VMALLOC_END 0xdf000000
#define VMALLOC_END 0xdf000000UL

2
arch/arm/mach-footbridge/include/mach/vmalloc.h
View file

@ -7,4 +7,4 @@
*/
#define VMALLOC_END 0xf0000000
#define VMALLOC_END 0xf0000000UL

2
arch/arm/mach-h720x/include/mach/vmalloc.h
View file

@ -5,6 +5,6 @@
#ifndef __ARCH_ARM_VMALLOC_H
#define __ARCH_ARM_VMALLOC_H
#define VMALLOC_END 0xd0000000
#define VMALLOC_END 0xd0000000UL
#endif

2
arch/arm/mach-imx/devices-imx25.h
View file

@ -79,7 +79,7 @@ imx25_sdhci_esdhc_imx_data[] __initconst;
#define imx25_add_sdhci_esdhc_imx(id, pdata) \
imx_add_sdhci_esdhc_imx(&imx25_sdhci_esdhc_imx_data[id], pdata)
extern const struct imx_spi_imx_data imx25_spi_imx_data[] __initconst;
extern const struct imx_spi_imx_data imx25_cspi_data[] __initconst;
#define imx25_add_spi_imx(id, pdata) \
imx_add_spi_imx(&imx25_spi_imx_data[id], pdata)
#define imx25_add_spi_imx0(pdata) imx25_add_spi_imx(0, pdata)

6
arch/arm/mach-imx/eukrea_mbimx27-baseboard.c
View file

@ -247,9 +247,6 @@ static const struct imxuart_platform_data uart_pdata __initconst = {
.flags = IMXUART_HAVE_RTSCTS,
};
#if defined(CONFIG_TOUCHSCREEN_ADS7846) \
|| defined(CONFIG_TOUCHSCREEN_ADS7846_MODULE)
#define ADS7846_PENDOWN (GPIO_PORTD | 25)
static void ads7846_dev_init(void)
@ -270,9 +267,7 @@ static struct ads7846_platform_data ads7846_config __initdata = {
.get_pendown_state = ads7846_get_pendown_state,
.keep_vref_on = 1,
};
#endif
#if defined(CONFIG_SPI_IMX) || defined(CONFIG_SPI_IMX_MODULE)
static struct spi_board_info eukrea_mbimx27_spi_board_info[] __initdata = {
[0] = {
.modalias = "ads7846",
@ -291,7 +286,6 @@ static const struct spi_imx_master eukrea_mbimx27_spi0_data __initconst = {
.chipselect = eukrea_mbimx27_spi_cs,
.num_chipselect = ARRAY_SIZE(eukrea_mbimx27_spi_cs),
};
#endif
static struct i2c_board_info eukrea_mbimx27_i2c_devices[] = {
{

2
arch/arm/mach-integrator/include/mach/vmalloc.h
View file

@ -17,4 +17,4 @@
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
#define VMALLOC_END 0xd0000000
#define VMALLOC_END 0xd0000000UL

2
arch/arm/mach-msm/include/mach/vmalloc.h
View file

@ -16,7 +16,7 @@
#ifndef __ASM_ARCH_MSM_VMALLOC_H
#define __ASM_ARCH_MSM_VMALLOC_H
#define VMALLOC_END 0xd0000000
#define VMALLOC_END 0xd0000000UL
#endif

3
arch/arm/mach-mx3/mach-pcm037_eet.c
View file

@ -14,6 +14,7 @@
#include <mach/common.h>
#include <mach/iomux-mx3.h>
#include <mach/spi.h>
#include <asm/mach-types.h>
@ -59,14 +60,12 @@ static struct spi_board_info pcm037_spi_dev[] = {
};
/* Platform Data for MXC CSPI */
#if defined(CONFIG_SPI_IMX) || defined(CONFIG_SPI_IMX_MODULE)
static int pcm037_spi1_cs[] = {MXC_SPI_CS(1), IOMUX_TO_GPIO(MX31_PIN_KEY_COL7)};
static const struct spi_imx_master pcm037_spi1_pdata __initconst = {
.chipselect = pcm037_spi1_cs,
.num_chipselect = ARRAY_SIZE(pcm037_spi1_cs),
};
#endif
/* GPIO-keys input device */
static struct gpio_keys_button pcm037_gpio_keys[] = {

2
arch/arm/mach-netx/include/mach/vmalloc.h
View file

@ -16,4 +16,4 @@
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
#define VMALLOC_END 0xd0000000
#define VMALLOC_END 0xd0000000UL

2
arch/arm/mach-omap1/include/mach/vmalloc.h
View file

@ -17,4 +17,4 @@
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
#define VMALLOC_END 0xd8000000
#define VMALLOC_END 0xd8000000UL

2
arch/arm/mach-omap2/include/mach/vmalloc.h
View file

@ -17,4 +17,4 @@
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
#define VMALLOC_END 0xf8000000
#define VMALLOC_END 0xf8000000UL

7
arch/arm/mach-omap2/pm24xx.c
View file

@ -30,6 +30,7 @@
#include <linux/irq.h>
#include <linux/time.h>
#include <linux/gpio.h>
#include <linux/console.h>
#include <asm/mach/time.h>
#include <asm/mach/irq.h>
@ -118,6 +119,10 @@ static void omap2_enter_full_retention(void)
if (omap_irq_pending())
goto no_sleep;
/* Block console output in case it is on one of the OMAP UARTs */
if (try_acquire_console_sem())
goto no_sleep;
omap_uart_prepare_idle(0);
omap_uart_prepare_idle(1);
omap_uart_prepare_idle(2);
@ -131,6 +136,8 @@ static void omap2_enter_full_retention(void)
omap_uart_resume_idle(1);
omap_uart_resume_idle(0);
release_console_sem();
no_sleep:
if (omap2_pm_debug) {
unsigned long long tmp;

10
arch/arm/mach-omap2/pm34xx.c
View file

@ -28,6 +28,7 @@
#include <linux/clk.h>
#include <linux/delay.h>
#include <linux/slab.h>
#include <linux/console.h>
#include <plat/sram.h>
#include <plat/clockdomain.h>
@ -385,6 +386,12 @@ void omap_sram_idle(void)
omap3_enable_io_chain();
}
/* Block console output in case it is on one of the OMAP UARTs */
if (per_next_state < PWRDM_POWER_ON ||
core_next_state < PWRDM_POWER_ON)
if (try_acquire_console_sem())
goto console_still_active;
/* PER */
if (per_next_state < PWRDM_POWER_ON) {
omap_uart_prepare_idle(2);
@ -463,6 +470,9 @@ void omap_sram_idle(void)
omap_uart_resume_idle(3);
}
release_console_sem();
console_still_active:
/* Disable IO-PAD and IO-CHAIN wakeup */
if (omap3_has_io_wakeup() &&
(per_next_state < PWRDM_POWER_ON ||

7
arch/arm/mach-omap2/serial.c
View file

@ -27,6 +27,7 @@
#include <linux/slab.h>
#include <linux/serial_8250.h>
#include <linux/pm_runtime.h>
#include <linux/console.h>
#ifdef CONFIG_SERIAL_OMAP
#include <plat/omap-serial.h>
@ -406,7 +407,7 @@ void omap_uart_resume_idle(int num)
struct omap_uart_state *uart;
list_for_each_entry(uart, &uart_list, node) {
if (num == uart->num) {
if (num == uart->num && uart->can_sleep) {
omap_uart_enable_clocks(uart);
/* Check for IO pad wakeup */
@ -807,6 +808,8 @@ void __init omap_serial_init_port(int port)
oh->dev_attr = uart;
acquire_console_sem(); /* in case the earlycon is on the UART */
/*
* Because of early UART probing, UART did not get idled
* on init. Now that omap_device is ready, ensure full idle
@ -831,6 +834,8 @@ void __init omap_serial_init_port(int port)
omap_uart_block_sleep(uart);
uart->timeout = DEFAULT_TIMEOUT;
release_console_sem();
if ((cpu_is_omap34xx() && uart->padconf) ||
(uart->wk_en && uart->wk_mask)) {
device_init_wakeup(&od->pdev.dev, true);

2
arch/arm/mach-pnx4008/include/mach/vmalloc.h
View file

@ -17,4 +17,4 @@
* The vmalloc() routines leaves a hole of 4kB between each vmalloced
* area for the same reason. ;)
*/
#define VMALLOC_END 0xd0000000
#define VMALLOC_END 0xd0000000UL

2
arch/arm/mach-rpc/include/mach/vmalloc.h
View file

@ -7,4 +7,4 @@
* it under the terms of the GNU General Public License version 2 as
* published by the Free Software Foundation.
*/
#define VMALLOC_END 0xdc000000
#define VMALLOC_END 0xdc000000UL

8
arch/arm/mach-s3c2410/h1940-bluetooth.c
View file

@ -77,13 +77,13 @@ static int __devinit h1940bt_probe(struct platform_device *pdev)
/* Configures BT serial port GPIOs */
s3c_gpio_cfgpin(S3C2410_GPH(0), S3C2410_GPH0_nCTS0);
s3c_gpio_cfgpull(S3C2410_GPH(0), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPH(0), S3C_GPIO_PULL_NONE);
s3c_gpio_cfgpin(S3C2410_GPH(1), S3C2410_GPIO_OUTPUT);
s3c_gpio_cfgpull(S3C2410_GPH(1), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPH(1), S3C_GPIO_PULL_NONE);
s3c_gpio_cfgpin(S3C2410_GPH(2), S3C2410_GPH2_TXD0);
s3c_gpio_cfgpull(S3C2410_GPH(2), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPH(2), S3C_GPIO_PULL_NONE);
s3c_gpio_cfgpin(S3C2410_GPH(3), S3C2410_GPH3_RXD0);
s3c_gpio_cfgpull(S3C2410_GPH(3), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPH(3), S3C_GPIO_PULL_NONE);
rfk = rfkill_alloc(DRV_NAME, &pdev->dev, RFKILL_TYPE_BLUETOOTH,

5
arch/arm/mach-s3c2416/irq.c
View file

@ -168,12 +168,11 @@ static struct irq_chip s3c2416_irq_dma = {
static void s3c2416_irq_demux_uart3(unsigned int irq, struct irq_desc *desc)
{
s3c2416_irq_demux(IRQ_S3C2443_UART3, 3);
s3c2416_irq_demux(IRQ_S3C2443_RX3, 3);
}
#define INTMSK_UART3 (1UL << (IRQ_S3C2443_UART3 - IRQ_EINT0))
#define SUBMSK_UART3 (0xf << (IRQ_S3C2443_RX3 - S3C2410_IRQSUB(0)))
#define SUBMSK_UART3 (0x7 << (IRQ_S3C2443_RX3 - S3C2410_IRQSUB(0)))
static void s3c2416_irq_uart3_mask(unsigned int irqno)
{

5
arch/arm/mach-s3c2443/irq.c
View file

@ -166,12 +166,11 @@ static struct irq_chip s3c2443_irq_dma = {
static void s3c2443_irq_demux_uart3(unsigned int irq, struct irq_desc *desc)
{
s3c2443_irq_demux(IRQ_S3C2443_UART3, 3);
s3c2443_irq_demux(IRQ_S3C2443_RX3, 3);
}
#define INTMSK_UART3 (1UL << (IRQ_S3C2443_UART3 - IRQ_EINT0))
#define SUBMSK_UART3 (0xf << (IRQ_S3C2443_RX3 - S3C2410_IRQSUB(0)))
#define SUBMSK_UART3 (0x7 << (IRQ_S3C2443_RX3 - S3C2410_IRQSUB(0)))
static void s3c2443_irq_uart3_mask(unsigned int irqno)
{

2
arch/arm/mach-s3c64xx/mach-mini6410.c
View file

@ -45,7 +45,7 @@
#include <video/platform_lcd.h>
#define UCON (S3C2410_UCON_DEFAULT | S3C2410_UCON_UCLK)
#define UCON S3C2410_UCON_DEFAULT
#define ULCON (S3C2410_LCON_CS8 | S3C2410_LCON_PNONE | S3C2410_LCON_STOPB)
#define UFCON (S3C2410_UFCON_RXTRIG8 | S3C2410_UFCON_FIFOMODE)

2
arch/arm/mach-s3c64xx/mach-real6410.c
View file

@ -46,7 +46,7 @@
#include <video/platform_lcd.h>
#define UCON (S3C2410_UCON_DEFAULT | S3C2410_UCON_UCLK)
#define UCON S3C2410_UCON_DEFAULT
#define ULCON (S3C2410_LCON_CS8 | S3C2410_LCON_PNONE | S3C2410_LCON_STOPB)
#define UFCON (S3C2410_UFCON_RXTRIG8 | S3C2410_UFCON_FIFOMODE)

1
arch/arm/mach-s5pv210/mach-smdkc110.c
View file

@ -13,6 +13,7 @@
#include <linux/init.h>
#include <linux/serial_core.h>
#include <linux/i2c.h>
#include <linux/sysdev.h>
#include <asm/mach/arch.h>
#include <asm/mach/map.h>

1
arch/arm/mach-s5pv210/mach-smdkv210.c
View file

@ -13,6 +13,7 @@
#include <linux/i2c.h>
#include <linux/init.h>
#include <linux/serial_core.h>
#include <linux/sysdev.h>
#include <asm/mach/arch.h>
#include <asm/mach/map.h>

2
arch/arm/mach-shark/include/mach/vmalloc.h
View file

@ -1,4 +1,4 @@
/*
* arch/arm/mach-shark/include/mach/vmalloc.h
*/
#define VMALLOC_END 0xd0000000
#define VMALLOC_END 0xd0000000UL

147
arch/arm/mach-shmobile/board-ap4evb.c
View file

@ -567,38 +567,127 @@ static struct platform_device *qhd_devices[] __initdata = {
/* FSI */
#define IRQ_FSI evt2irq(0x1840)
static int __fsi_set_rate(struct clk *clk, long rate, int enable)
{
int ret = 0;
static int fsi_set_rate(int is_porta, int rate)
if (rate <= 0)
return ret;
if (enable) {
ret = clk_set_rate(clk, rate);
if (0 == ret)
ret = clk_enable(clk);
} else {
clk_disable(clk);
}
return ret;
}
static int __fsi_set_round_rate(struct clk *clk, long rate, int enable)
{
return __fsi_set_rate(clk, clk_round_rate(clk, rate), enable);
}
static int fsi_ak4642_set_rate(struct device *dev, int rate, int enable)
{
struct clk *fsia_ick;
struct clk *fsiack;
int ret = -EIO;
fsia_ick = clk_get(dev, "icka");
if (IS_ERR(fsia_ick))
return PTR_ERR(fsia_ick);
/*
* FSIACK is connected to AK4642,
* and use external clock pin from it.
* it is parent of fsia_ick now.
*/
fsiack = clk_get_parent(fsia_ick);
if (!fsiack)
goto fsia_ick_out;
/*
* we get 1/1 divided clock by setting same rate to fsiack and fsia_ick
*
** FIXME **
* Because the freq_table of external clk (fsiack) are all 0,
* the return value of clk_round_rate became 0.
* So, it use __fsi_set_rate here.
*/
ret = __fsi_set_rate(fsiack, rate, enable);
if (ret < 0)
goto fsiack_out;
ret = __fsi_set_round_rate(fsia_ick, rate, enable);
if ((ret < 0) && enable)
__fsi_set_round_rate(fsiack, rate, 0); /* disable FSI ACK */
fsiack_out:
clk_put(fsiack);
fsia_ick_out:
clk_put(fsia_ick);
return 0;
}
static int fsi_hdmi_set_rate(struct device *dev, int rate, int enable)
{
struct clk *fsib_clk;
struct clk *fdiv_clk = &sh7372_fsidivb_clk;
long fsib_rate = 0;
long fdiv_rate = 0;
int ackmd_bpfmd;
int ret;
/* set_rate is not needed if port A */
if (is_porta)
return 0;
fsib_clk = clk_get(NULL, "fsib_clk");
if (IS_ERR(fsib_clk))
return -EINVAL;
switch (rate) {
case 44100:
clk_set_rate(fsib_clk, clk_round_rate(fsib_clk, 11283000));
ret = SH_FSI_ACKMD_256 | SH_FSI_BPFMD_64;
fsib_rate = rate * 256;
ackmd_bpfmd = SH_FSI_ACKMD_256 | SH_FSI_BPFMD_64;
break;
case 48000:
clk_set_rate(fsib_clk, clk_round_rate(fsib_clk, 85428000));
clk_set_rate(fdiv_clk, clk_round_rate(fdiv_clk, 12204000));
ret = SH_FSI_ACKMD_256 | SH_FSI_BPFMD_64;
fsib_rate = 85428000; /* around 48kHz x 256 x 7 */
fdiv_rate = rate * 256;
ackmd_bpfmd = SH_FSI_ACKMD_256 | SH_FSI_BPFMD_64;
break;
default:
pr_err("unsupported rate in FSI2 port B\n");
ret = -EINVAL;
break;
return -EINVAL;
}
/* FSI B setting */
fsib_clk = clk_get(dev, "ickb");
if (IS_ERR(fsib_clk))
return -EIO;
ret = __fsi_set_round_rate(fsib_clk, fsib_rate, enable);
clk_put(fsib_clk);
if (ret < 0)
return ret;
/* FSI DIV setting */
ret = __fsi_set_round_rate(fdiv_clk, fdiv_rate, enable);
if (ret < 0) {
/* disable FSI B */
if (enable)
__fsi_set_round_rate(fsib_clk, fsib_rate, 0);
return ret;
}
return ackmd_bpfmd;
}
static int fsi_set_rate(struct device *dev, int is_porta, int rate, int enable)
{
int ret;
if (is_porta)
ret = fsi_ak4642_set_rate(dev, rate, enable);
else
ret = fsi_hdmi_set_rate(dev, rate, enable);
return ret;
}
@ -880,6 +969,11 @@ static int __init hdmi_init_pm_clock(void)
goto out;
}
ret = clk_enable(&sh7372_pllc2_clk);
if (ret < 0) {
pr_err("Cannot enable pllc2 clock\n");
goto out;
}
pr_debug("PLLC2 set frequency %lu\n", rate);
ret = clk_set_parent(hdmi_ick, &sh7372_pllc2_clk);
@ -896,23 +990,11 @@ static int __init hdmi_init_pm_clock(void)
device_initcall(hdmi_init_pm_clock);
#define FSIACK_DUMMY_RATE 48000
static int __init fsi_init_pm_clock(void)
{
struct clk *fsia_ick;
int ret;
/*
* FSIACK is connected to AK4642,
* and the rate is depend on playing sound rate.
* So, set dummy rate (= 48k) here
*/
ret = clk_set_rate(&sh7372_fsiack_clk, FSIACK_DUMMY_RATE);
if (ret < 0) {
pr_err("Cannot set FSIACK dummy rate: %d\n", ret);
return ret;
}
fsia_ick = clk_get(&fsi_device.dev, "icka");
if (IS_ERR(fsia_ick)) {
ret = PTR_ERR(fsia_ick);
@ -921,16 +1003,9 @@ static int __init fsi_init_pm_clock(void)
}
ret = clk_set_parent(fsia_ick, &sh7372_fsiack_clk);
if (ret < 0) {
pr_err("Cannot set FSI-A parent: %d\n", ret);
goto out;
}
ret = clk_set_rate(fsia_ick, FSIACK_DUMMY_RATE);
if (ret < 0)
pr_err("Cannot set FSI-A rate: %d\n", ret);
pr_err("Cannot set FSI-A parent: %d\n", ret);
out:
clk_put(fsia_ick);
return ret;

39
arch/arm/mach-shmobile/clock-sh7372.c
View file

@ -220,8 +220,7 @@ static void pllc2_disable(struct clk *clk)
__raw_writel(__raw_readl(PLLC2CR) & ~0x80000000, PLLC2CR);
}
static int pllc2_set_rate(struct clk *clk,
unsigned long rate, int algo_id)
static int pllc2_set_rate(struct clk *clk, unsigned long rate)
{
unsigned long value;
int idx;
@ -230,21 +229,13 @@ static int pllc2_set_rate(struct clk *clk,
if (idx < 0)
return idx;
if (rate == clk->parent->rate) {
pllc2_disable(clk);
return 0;
}
if (rate == clk->parent->rate)
return -EINVAL;
value = __raw_readl(PLLC2CR) & ~(0x3f << 24);
if (value & 0x80000000)
pllc2_disable(clk);
__raw_writel((value & ~0x80000000) | ((idx + 19) << 24), PLLC2CR);
if (value & 0x80000000)
return pllc2_enable(clk);
return 0;
}
@ -453,32 +444,24 @@ static int fsidiv_enable(struct clk *clk)
unsigned long value;
value = __raw_readl(clk->mapping->base) >> 16;
if (value < 2) {
fsidiv_disable(clk);
return -ENOENT;
}
if (value < 2)
return -EIO;
__raw_writel((value << 16) | 0x3, clk->mapping->base);
return 0;
}
static int fsidiv_set_rate(struct clk *clk,
unsigned long rate, int algo_id)
static int fsidiv_set_rate(struct clk *clk, unsigned long rate)
{
int idx;
if (clk->parent->rate == rate) {
fsidiv_disable(clk);
return 0;
}
idx = (clk->parent->rate / rate) & 0xffff;
if (idx < 2)
return -ENOENT;
return -EINVAL;
__raw_writel(idx << 16, clk->mapping->base);
return fsidiv_enable(clk);
return 0;
}
static struct clk_ops fsidiv_clk_ops = {
@ -609,8 +592,6 @@ static struct clk_lookup lookups[] = {
CLKDEV_CON_ID("vck3_clk", &div6_clks[DIV6_VCK3]),
CLKDEV_CON_ID("fmsi_clk", &div6_clks[DIV6_FMSI]),
CLKDEV_CON_ID("fmso_clk", &div6_clks[DIV6_FMSO]),
CLKDEV_CON_ID("fsia_clk", &div6_reparent_clks[DIV6_FSIA]),
CLKDEV_CON_ID("fsib_clk", &div6_reparent_clks[DIV6_FSIB]),
CLKDEV_CON_ID("sub_clk", &div6_clks[DIV6_SUB]),
CLKDEV_CON_ID("spu_clk", &div6_clks[DIV6_SPU]),
CLKDEV_CON_ID("vou_clk", &div6_clks[DIV6_VOU]),
@ -647,8 +628,8 @@ static struct clk_lookup lookups[] = {
CLKDEV_DEV_ID("sh_cmt.10", &mstp_clks[MSTP329]), /* CMT10 */
CLKDEV_DEV_ID("sh_fsi2", &mstp_clks[MSTP328]), /* FSI2 */
CLKDEV_DEV_ID("i2c-sh_mobile.1", &mstp_clks[MSTP323]), /* IIC1 */
CLKDEV_DEV_ID("r8a66597_hcd.0", &mstp_clks[MSTP323]), /* USB0 */
CLKDEV_DEV_ID("r8a66597_udc.0", &mstp_clks[MSTP323]), /* USB0 */
CLKDEV_DEV_ID("r8a66597_hcd.0", &mstp_clks[MSTP322]), /* USB0 */
CLKDEV_DEV_ID("r8a66597_udc.0", &mstp_clks[MSTP322]), /* USB0 */
CLKDEV_DEV_ID("sh_mobile_sdhi.0", &mstp_clks[MSTP314]), /* SDHI0 */
CLKDEV_DEV_ID("sh_mobile_sdhi.1", &mstp_clks[MSTP313]), /* SDHI1 */
CLKDEV_DEV_ID("sh_mmcif.0", &mstp_clks[MSTP312]), /* MMC */

6
arch/arm/mach-ux500/cpu.c
View file

@ -75,14 +75,14 @@ void __init ux500_init_irq(void)
static inline void ux500_cache_wait(void __iomem *reg, unsigned long mask)
{
/* wait for the operation to complete */
while (readl(reg) & mask)
while (readl_relaxed(reg) & mask)
;
}
static inline void ux500_cache_sync(void)
{
void __iomem *base = __io_address(UX500_L2CC_BASE);
writel(0, base + L2X0_CACHE_SYNC);
writel_relaxed(0, base + L2X0_CACHE_SYNC);
ux500_cache_wait(base + L2X0_CACHE_SYNC, 1);
}
@ -107,7 +107,7 @@ static void ux500_l2x0_inv_all(void)
uint32_t l2x0_way_mask = (1<<16) - 1; /* Bitmask of active ways */
/* invalidate all ways */
writel(l2x0_way_mask, l2x0_base + L2X0_INV_WAY);
writel_relaxed(l2x0_way_mask, l2x0_base + L2X0_INV_WAY);
ux500_cache_wait(l2x0_base + L2X0_INV_WAY, l2x0_way_mask);
ux500_cache_sync();
}

2
arch/arm/mach-versatile/include/mach/vmalloc.h
View file

@ -18,4 +18,4 @@
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
#define VMALLOC_END 0xd8000000
#define VMALLOC_END 0xd8000000UL

4
arch/arm/mm/ioremap.c
View file

@ -206,8 +206,8 @@ void __iomem * __arm_ioremap_pfn_caller(unsigned long pfn,
*/
if (pfn_valid(pfn)) {
printk(KERN_WARNING "BUG: Your driver calls ioremap() on system memory. This leads\n"
KERN_WARNING "to architecturally unpredictable behaviour on ARMv6+, and ioremap()\n"
KERN_WARNING "will fail in the next kernel release. Please fix your driver.\n");
"to architecturally unpredictable behaviour on ARMv6+, and ioremap()\n"
"will fail in the next kernel release. Please fix your driver.\n");
WARN_ON(1);
}

8
arch/arm/plat-mxc/devices/platform-imx-dma.c
View file

@ -12,15 +12,7 @@
#include <mach/hardware.h>
#include <mach/devices-common.h>
#ifdef SDMA_IS_MERGED
#include <mach/sdma.h>
#else
struct sdma_platform_data {
int sdma_version;
char *cpu_name;
int to_version;
};
#endif
struct imx_imx_sdma_data {
resource_size_t iobase;

1
arch/arm/plat-mxc/devices/platform-spi_imx.c
View file

@ -27,6 +27,7 @@ const struct imx_spi_imx_data imx21_cspi_data[] __initconst = {
imx_spi_imx_data_entry(MX21, CSPI, "imx21-cspi", _id, _hwid, SZ_4K)
imx21_cspi_data_entry(0, 1),
imx21_cspi_data_entry(1, 2),
};
#endif
#ifdef CONFIG_SOC_IMX25

89
arch/arm/plat-nomadik/timer.c
View file

@ -3,6 +3,7 @@
*
* Copyright (C) 2008 STMicroelectronics
* Copyright (C) 2010 Alessandro Rubini
* Copyright (C) 2010 Linus Walleij for ST-Ericsson
*
* This program is free software; you can redistribute it and/or modify
* it under the terms of the GNU General Public License version 2, as
@ -16,11 +17,13 @@
#include <linux/clk.h>
#include <linux/jiffies.h>
#include <linux/err.h>
#include <linux/cnt32_to_63.h>
#include <linux/timer.h>
#include <asm/mach/time.h>
#include <plat/mtu.h>
void __iomem *mtu_base; /* ssigned by machine code */
void __iomem *mtu_base; /* Assigned by machine code */
/*
* Kernel assumes that sched_clock can be called early
@ -48,16 +51,82 @@ static struct clocksource nmdk_clksrc = {
/*
* Override the global weak sched_clock symbol with this
* local implementation which uses the clocksource to get some
* better resolution when scheduling the kernel. We accept that
* this wraps around for now, since it is just a relative time
* stamp. (Inspired by OMAP implementation.)
* better resolution when scheduling the kernel.
*
* Because the hardware timer period may be quite short
* (32.3 secs on the 133 MHz MTU timer selection on ux500)
* and because cnt32_to_63() needs to be called at least once per
* half period to work properly, a kernel keepwarm() timer is set up
* to ensure this requirement is always met.
*
* Also the sched_clock timer will wrap around at some point,
* here we set it to run continously for a year.
*/
#define SCHED_CLOCK_MIN_WRAP 3600*24*365
static struct timer_list cnt32_to_63_keepwarm_timer;
static u32 sched_mult;
static u32 sched_shift;
unsigned long long notrace sched_clock(void)
{
return clocksource_cyc2ns(nmdk_clksrc.read(
&nmdk_clksrc),
nmdk_clksrc.mult,
nmdk_clksrc.shift);
u64 cycles;
if (unlikely(!mtu_base))
return 0;
cycles = cnt32_to_63(-readl(mtu_base + MTU_VAL(0)));
/*
* sched_mult is guaranteed to be even so will
* shift out bit 63
*/
return (cycles * sched_mult) >> sched_shift;
}
/* Just kick sched_clock every so often */
static void cnt32_to_63_keepwarm(unsigned long data)
{
mod_timer(&cnt32_to_63_keepwarm_timer, round_jiffies(jiffies + data));
(void) sched_clock();
}
/*
* Set up a timer to keep sched_clock():s 32_to_63 algorithm warm
* once in half a 32bit timer wrap interval.
*/
static void __init nmdk_sched_clock_init(unsigned long rate)
{
u32 v;
unsigned long delta;
u64 days;
/* Find the apropriate mult and shift factors */
clocks_calc_mult_shift(&sched_mult, &sched_shift,
rate, NSEC_PER_SEC, SCHED_CLOCK_MIN_WRAP);
/* We need to multiply by an even number to get rid of bit 63 */
if (sched_mult & 1)
sched_mult++;
/* Let's see what we get, take max counter and scale it */
days = (0xFFFFFFFFFFFFFFFFLLU * sched_mult) >> sched_shift;
do_div(days, NSEC_PER_SEC);
do_div(days, (3600*24));
pr_info("sched_clock: using %d bits @ %lu Hz wrap in %lu days\n",
(64 - sched_shift), rate, (unsigned long) days);
/*
* Program a timer to kick us at half 32bit wraparound
* Formula: seconds per wrap = (2^32) / f
*/
v = 0xFFFFFFFFUL / rate;
/* We want half of the wrap time to keep cnt32_to_63 warm */
v /= 2;
pr_debug("sched_clock: prescaled timer rate: %lu Hz, "
"initialize keepwarm timer every %d seconds\n", rate, v);
/* Convert seconds to jiffies */
delta = msecs_to_jiffies(v*1000);
setup_timer(&cnt32_to_63_keepwarm_timer, cnt32_to_63_keepwarm, delta);
mod_timer(&cnt32_to_63_keepwarm_timer, round_jiffies(jiffies + delta));
}
/* Clockevent device: use one-shot mode */
@ -161,13 +230,15 @@ void __init nmdk_timer_init(void)
writel(0, mtu_base + MTU_BGLR(0));
writel(cr | MTU_CRn_ENA, mtu_base + MTU_CR(0));
/* Now the scheduling clock is ready */
/* Now the clock source is ready */
nmdk_clksrc.read = nmdk_read_timer;
if (clocksource_register(&nmdk_clksrc))
pr_err("timer: failed to initialize clock source %s\n",
nmdk_clksrc.name);
nmdk_sched_clock_init(rate);
/* Timer 1 is used for events */
clockevents_calc_mult_shift(&nmdk_clkevt, rate, MTU_MIN_RANGE);

3
arch/arm/plat-pxa/include/plat/sdhci.h
View file

@ -17,6 +17,9 @@
/* Require clock free running */
#define PXA_FLAG_DISABLE_CLOCK_GATING (1<<0)
/* Board design supports 8-bit data on SD/SDIO BUS */
#define PXA_FLAG_SD_8_BIT_CAPABLE_SLOT (1<<2)
/*
* struct pxa_sdhci_platdata() - Platform device data for PXA SDHCI
* @max_speed: the maximum speed supported

6
arch/arm/plat-s3c24xx/spi-bus0-gpe11_12_13.c
View file

@ -29,8 +29,8 @@ void s3c24xx_spi_gpiocfg_bus0_gpe11_12_13(struct s3c2410_spi_info *spi,
} else {
s3c_gpio_cfgpin(S3C2410_GPE(13), S3C2410_GPIO_INPUT);
s3c_gpio_cfgpin(S3C2410_GPE(11), S3C2410_GPIO_INPUT);
s3c_gpio_cfgpull(S3C2410_GPE(11), S3C_GPIO_PULL_NONE);
s3c_gpio_cfgpull(S3C2410_GPE(12), S3C_GPIO_PULL_NONE);
s3c_gpio_cfgpull(S3C2410_GPE(13), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPE(11), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPE(12), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPE(13), S3C_GPIO_PULL_NONE);
}
}

6
arch/arm/plat-s3c24xx/spi-bus1-gpd8_9_10.c
View file

@ -31,8 +31,8 @@ void s3c24xx_spi_gpiocfg_bus1_gpd8_9_10(struct s3c2410_spi_info *spi,
} else {
s3c_gpio_cfgpin(S3C2410_GPD(8), S3C2410_GPIO_INPUT);
s3c_gpio_cfgpin(S3C2410_GPD(9), S3C2410_GPIO_INPUT);
s3c_gpio_cfgpull(S3C2410_GPD(10), S3C_GPIO_PULL_NONE);
s3c_gpio_cfgpull(S3C2410_GPD(9), S3C_GPIO_PULL_NONE);
s3c_gpio_cfgpull(S3C2410_GPD(8), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPD(10), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPD(9), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPD(8), S3C_GPIO_PULL_NONE);
}
}

6
arch/arm/plat-s3c24xx/spi-bus1-gpg5_6_7.c
View file

@ -29,8 +29,8 @@ void s3c24xx_spi_gpiocfg_bus1_gpg5_6_7(struct s3c2410_spi_info *spi,
} else {
s3c_gpio_cfgpin(S3C2410_GPG(7), S3C2410_GPIO_INPUT);
s3c_gpio_cfgpin(S3C2410_GPG(5), S3C2410_GPIO_INPUT);
s3c_gpio_cfgpull(S3C2410_GPG(5), S3C_GPIO_PULL_NONE);
s3c_gpio_cfgpull(S3C2410_GPG(6), S3C_GPIO_PULL_NONE);
s3c_gpio_cfgpull(S3C2410_GPG(7), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPG(5), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPG(6), S3C_GPIO_PULL_NONE);
s3c_gpio_setpull(S3C2410_GPG(7), S3C_GPIO_PULL_NONE);
}
}

2
arch/powerpc/mm/pgtable.c
View file

@ -92,7 +92,7 @@ static void pte_free_rcu_callback(struct rcu_head *head)
static void pte_free_submit(struct pte_freelist_batch *batch)
{
call_rcu(&batch->rcu, pte_free_rcu_callback);
call_rcu_sched(&batch->rcu, pte_free_rcu_callback);
}
void pgtable_free_tlb(struct mmu_gather *tlb, void *table, unsigned shift)

10
arch/s390/kernel/nmi.c
View file

@ -95,7 +95,6 @@ EXPORT_SYMBOL_GPL(s390_handle_mcck);
static int notrace s390_revalidate_registers(struct mci *mci)
{
int kill_task;
u64 tmpclock;
u64 zero;
void *fpt_save_area, *fpt_creg_save_area;
@ -214,11 +213,10 @@ static int notrace s390_revalidate_registers(struct mci *mci)
: "0", "cc");
#endif
/* Revalidate clock comparator register */
asm volatile(
" stck 0(%1)\n"
" sckc 0(%1)"
: "=m" (tmpclock) : "a" (&(tmpclock)) : "cc", "memory");
if (S390_lowcore.clock_comparator == -1)
set_clock_comparator(S390_lowcore.mcck_clock);
else
set_clock_comparator(S390_lowcore.clock_comparator);
/* Check if old PSW is valid */
if (!mci->wp)
/*

14
arch/s390/lib/delay.c
View file

@ -29,17 +29,21 @@ static void __udelay_disabled(unsigned long long usecs)
{
unsigned long mask, cr0, cr0_saved;
u64 clock_saved;
u64 end;
mask = psw_kernel_bits | PSW_MASK_WAIT | PSW_MASK_EXT;
end = get_clock() + (usecs << 12);
clock_saved = local_tick_disable();
set_clock_comparator(get_clock() + (usecs << 12));
__ctl_store(cr0_saved, 0, 0);
cr0 = (cr0_saved & 0xffff00e0) | 0x00000800;
__ctl_load(cr0 , 0, 0);
mask = psw_kernel_bits | PSW_MASK_WAIT | PSW_MASK_EXT;
lockdep_off();
trace_hardirqs_on();
__load_psw_mask(mask);
local_irq_disable();
do {
set_clock_comparator(end);
trace_hardirqs_on();
__load_psw_mask(mask);
local_irq_disable();
} while (get_clock() < end);
lockdep_on();
__ctl_load(cr0_saved, 0, 0);
local_tick_enable(clock_saved);

7
arch/sh/include/asm/processor_32.h
View file

@ -199,10 +199,13 @@ extern unsigned long get_wchan(struct task_struct *p);
#define ARCH_HAS_PREFETCHW
static inline void prefetch(void *x)
{
__asm__ __volatile__ ("pref @%0\n\t" : : "r" (x) : "memory");
__builtin_prefetch(x, 0, 3);
}
#define prefetchw(x) prefetch(x)
static inline void prefetchw(void *x)
{
__builtin_prefetch(x, 1, 3);
}
#endif
#endif /* __KERNEL__ */

2
arch/sh/kernel/cpu/sh4/clock-sh4-202.c
View file

@ -110,7 +110,7 @@ static int shoc_clk_verify_rate(struct clk *clk, unsigned long rate)
return 0;
}
static int shoc_clk_set_rate(struct clk *clk, unsigned long rate, int algo_id)
static int shoc_clk_set_rate(struct clk *clk, unsigned long rate)
{
unsigned long frqcr3;
unsigned int tmp;

2
arch/sh/kernel/sys_sh.c
View file

@ -88,7 +88,7 @@ asmlinkage int sys_cacheflush(unsigned long addr, unsigned long len, int op)
}
if (op & CACHEFLUSH_I)
flush_cache_all();
flush_icache_range(addr, addr+len);
up_read(&current->mm->mmap_sem);
return 0;

2
arch/sh/kernel/vsyscall/vsyscall-trapa.S
View file

@ -8,9 +8,9 @@ __kernel_vsyscall:
* fill out .eh_frame -- PFM. */
.LEND_vsyscall:
.size __kernel_vsyscall,.-.LSTART_vsyscall
.previous
.section .eh_frame,"a",@progbits
.previous
.LCIE:
.ualong .LCIE_end - .LCIE_start
.LCIE_start:

12
arch/tile/Kconfig
View file

@ -329,6 +329,18 @@ endmenu # Tilera-specific configuration
menu "Bus options"
config PCI
bool "PCI support"
default y
select PCI_DOMAINS
---help---
Enable PCI root complex support, so PCIe endpoint devices can
be attached to the Tile chip. Many, but not all, PCI devices
are supported under Tilera's root complex driver.
config PCI_DOMAINS
bool
config NO_IOMEM
def_bool !PCI

52
arch/tile/include/asm/cacheflush.h
View file

@ -137,4 +137,56 @@ static inline void finv_buffer(void *buffer, size_t size)
mb_incoherent();
}
/*
* Flush & invalidate a VA range that is homed remotely on a single core,
* waiting until the memory controller holds the flushed values.
*/
static inline void finv_buffer_remote(void *buffer, size_t size)
{
char *p;
int i;
/*
* Flush and invalidate the buffer out of the local L1/L2
* and request the home cache to flush and invalidate as well.
*/
__finv_buffer(buffer, size);
/*
* Wait for the home cache to acknowledge that it has processed
* all the flush-and-invalidate requests. This does not mean
* that the flushed data has reached the memory controller yet,
* but it does mean the home cache is processing the flushes.
*/
__insn_mf();
/*
* Issue a load to the last cache line, which can't complete
* until all the previously-issued flushes to the same memory
* controller have also completed. If we weren't striping
* memory, that one load would be sufficient, but since we may
* be, we also need to back up to the last load issued to
* another memory controller, which would be the point where
* we crossed an 8KB boundary (the granularity of striping
* across memory controllers). Keep backing up and doing this
* until we are before the beginning of the buffer, or have
* hit all the controllers.
*/
for (i = 0, p = (char *)buffer + size - 1;
i < (1 << CHIP_LOG_NUM_MSHIMS()) && p >= (char *)buffer;
++i) {
const unsigned long STRIPE_WIDTH = 8192;
/* Force a load instruction to issue. */
*(volatile char *)p;
/* Jump to end of previous stripe. */
p -= STRIPE_WIDTH;
p = (char *)((unsigned long)p | (STRIPE_WIDTH - 1));
}
/* Wait for the loads (and thus flushes) to have completed. */
__insn_mf();
}
#endif /* _ASM_TILE_CACHEFLUSH_H */

15
arch/tile/include/asm/io.h
View file

@ -55,9 +55,6 @@ extern void iounmap(volatile void __iomem *addr);
#define ioremap_writethrough(physaddr, size) ioremap(physaddr, size)
#define ioremap_fullcache(physaddr, size) ioremap(physaddr, size)
void __iomem *ioport_map(unsigned long port, unsigned int len);
extern inline void ioport_unmap(void __iomem *addr) {}
#define mmiowb()
/* Conversion between virtual and physical mappings. */
@ -189,12 +186,22 @@ static inline void memcpy_toio(volatile void __iomem *dst, const void *src,
* we never run, uses them unconditionally.
*/
static inline int ioport_panic(void)
static inline long ioport_panic(void)
{
panic("inb/outb and friends do not exist on tile");
return 0;
}
static inline void __iomem *ioport_map(unsigned long port, unsigned int len)
{
return (void __iomem *) ioport_panic();
}
static inline void ioport_unmap(void __iomem *addr)
{
ioport_panic();
}
static inline u8 inb(unsigned long addr)
{
return ioport_panic();

117
arch/tile/include/asm/pci-bridge.h
View file

@ -1,117 +0,0 @@
/*
* Copyright 2010 Tilera Corporation. All Rights Reserved.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation, version 2.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
* NON INFRINGEMENT. See the GNU General Public License for
* more details.
*/
#ifndef _ASM_TILE_PCI_BRIDGE_H
#define _ASM_TILE_PCI_BRIDGE_H
#include <linux/ioport.h>
#include <linux/pci.h>
struct device_node;
struct pci_controller;
/*
* pci_io_base returns the memory address at which you can access
* the I/O space for PCI bus number `bus' (or NULL on error).
*/
extern void __iomem *pci_bus_io_base(unsigned int bus);
extern unsigned long pci_bus_io_base_phys(unsigned int bus);
extern unsigned long pci_bus_mem_base_phys(unsigned int bus);
/* Allocate a new PCI host bridge structure */
extern struct pci_controller *pcibios_alloc_controller(void);
/* Helper function for setting up resources */
extern void pci_init_resource(struct resource *res, unsigned long start,
unsigned long end, int flags, char *name);
/* Get the PCI host controller for a bus */
extern struct pci_controller *pci_bus_to_hose(int bus);
/*
* Structure of a PCI controller (host bridge)
*/
struct pci_controller {
int index; /* PCI domain number */
struct pci_bus *root_bus;
int first_busno;
int last_busno;
int hv_cfg_fd[2]; /* config{0,1} fds for this PCIe controller */
int hv_mem_fd; /* fd to Hypervisor for MMIO operations */
struct pci_ops *ops;
int irq_base; /* Base IRQ from the Hypervisor */
int plx_gen1; /* flag for PLX Gen 1 configuration */
/* Address ranges that are routed to this controller/bridge. */
struct resource mem_resources[3];
};
static inline struct pci_controller *pci_bus_to_host(struct pci_bus *bus)
{
return bus->sysdata;
}
extern void setup_indirect_pci_nomap(struct pci_controller *hose,
void __iomem *cfg_addr, void __iomem *cfg_data);
extern void setup_indirect_pci(struct pci_controller *hose,
u32 cfg_addr, u32 cfg_data);
extern void setup_grackle(struct pci_controller *hose);
extern unsigned char common_swizzle(struct pci_dev *, unsigned char *);
/*
* The following code swizzles for exactly one bridge. The routine
* common_swizzle below handles multiple bridges. But there are a
* some boards that don't follow the PCI spec's suggestion so we
* break this piece out separately.
*/
static inline unsigned char bridge_swizzle(unsigned char pin,
unsigned char idsel)
{
return (((pin-1) + idsel) % 4) + 1;
}
/*
* The following macro is used to lookup irqs in a standard table
* format for those PPC systems that do not already have PCI
* interrupts properly routed.
*/
/* FIXME - double check this */
#define PCI_IRQ_TABLE_LOOKUP ({ \
long _ctl_ = -1; \
if (idsel >= min_idsel && idsel <= max_idsel && pin <= irqs_per_slot) \
_ctl_ = pci_irq_table[idsel - min_idsel][pin-1]; \
_ctl_; \
})
/*
* Scan the buses below a given PCI host bridge and assign suitable
* resources to all devices found.
*/
extern int pciauto_bus_scan(struct pci_controller *, int);
#ifdef CONFIG_PCI
extern unsigned long pci_address_to_pio(phys_addr_t address);
#else
static inline unsigned long pci_address_to_pio(phys_addr_t address)
{
return (unsigned long)-1;
}
#endif
#endif /* _ASM_TILE_PCI_BRIDGE_H */

107
arch/tile/include/asm/pci.h
View file

@ -15,7 +15,29 @@
#ifndef _ASM_TILE_PCI_H
#define _ASM_TILE_PCI_H
#include <asm/pci-bridge.h>
#include <linux/pci.h>
/*
* Structure of a PCI controller (host bridge)
*/
struct pci_controller {
int index; /* PCI domain number */
struct pci_bus *root_bus;
int first_busno;
int last_busno;
int hv_cfg_fd[2]; /* config{0,1} fds for this PCIe controller */
int hv_mem_fd; /* fd to Hypervisor for MMIO operations */
struct pci_ops *ops;
int irq_base; /* Base IRQ from the Hypervisor */
int plx_gen1; /* flag for PLX Gen 1 configuration */
/* Address ranges that are routed to this controller/bridge. */
struct resource mem_resources[3];
};
/*
* The hypervisor maps the entirety of CPA-space as bus addresses, so
@ -24,57 +46,13 @@
*/
#define PCI_DMA_BUS_IS_PHYS 1
struct pci_controller *pci_bus_to_hose(int bus);
unsigned char __init common_swizzle(struct pci_dev *dev, unsigned char *pinp);
int __init tile_pci_init(void);
void pci_iounmap(struct pci_dev *dev, void __iomem *addr);
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long max);
static inline void pci_iounmap(struct pci_dev *dev, void __iomem *addr) {}
void __devinit pcibios_fixup_bus(struct pci_bus *bus);
int __devinit _tile_cfg_read(struct pci_controller *hose,
int bus,
int slot,
int function,
int offset,
int size,
u32 *val);
int __devinit _tile_cfg_write(struct pci_controller *hose,
int bus,
int slot,
int function,
int offset,
int size,
u32 val);
/*
* These are used to to config reads and writes in the early stages of
* setup before the driver infrastructure has been set up enough to be
* able to do config reads and writes.
*/
#define early_cfg_read(where, size, value) \
_tile_cfg_read(controller, \
current_bus, \
pci_slot, \
pci_fn, \
where, \
size, \
value)
#define early_cfg_write(where, size, value) \
_tile_cfg_write(controller, \
current_bus, \
pci_slot, \
pci_fn, \
where, \
size, \
value)
#define PCICFG_BYTE 1
#define PCICFG_WORD 2
#define PCICFG_DWORD 4
#define TILE_NUM_PCIE 2
#define pci_domain_nr(bus) (((struct pci_controller *)(bus)->sysdata)->index)
@ -88,33 +66,33 @@ static inline int pci_proc_domain(struct pci_bus *bus)
}
/*
* I/O space is currently not supported.
* pcibios_assign_all_busses() tells whether or not the bus numbers
* should be reassigned, in case the BIOS didn't do it correctly, or
* in case we don't have a BIOS and we want to let Linux do it.
*/
static inline int pcibios_assign_all_busses(void)
{
return 1;
}
#define TILE_PCIE_LOWER_IO 0x0
#define TILE_PCIE_UPPER_IO 0x10000
#define TILE_PCIE_PCIE_IO_SIZE 0x0000FFFF
#define _PAGE_NO_CACHE 0
#define _PAGE_GUARDED 0
#define pcibios_assign_all_busses() pci_assign_all_buses
extern int pci_assign_all_buses;
/*
* No special bus mastering setup handling.
*/
static inline void pcibios_set_master(struct pci_dev *dev)
{
/* No special bus mastering setup handling */
}
#define PCIBIOS_MIN_MEM 0
#define PCIBIOS_MIN_IO TILE_PCIE_LOWER_IO
#define PCIBIOS_MIN_IO 0
/*
* This flag tells if the platform is TILEmpower that needs
* special configuration for the PLX switch chip.
*/
extern int blade_pci;
extern int tile_plx_gen1;
/* Use any cpu for PCI. */
#define cpumask_of_pcibus(bus) cpu_online_mask
/* implement the pci_ DMA API in terms of the generic device dma_ one */
#include <asm-generic/pci-dma-compat.h>
@ -122,7 +100,4 @@ extern int blade_pci;
/* generic pci stuff */
#include <asm-generic/pci.h>
/* Use any cpu for PCI. */
#define cpumask_of_pcibus(bus) cpu_online_mask
#endif /* _ASM_TILE_PCI_H */

10
arch/tile/include/asm/processor.h
View file

@ -292,8 +292,18 @@ extern int kstack_hash;
/* Are we using huge pages in the TLB for kernel data? */
extern int kdata_huge;
/* Support standard Linux prefetching. */
#define ARCH_HAS_PREFETCH
#define prefetch(x) __builtin_prefetch(x)
#define PREFETCH_STRIDE CHIP_L2_LINE_SIZE()
/* Bring a value into the L1D, faulting the TLB if necessary. */
#ifdef __tilegx__
#define prefetch_L1(x) __insn_prefetch_l1_fault((void *)(x))
#else
#define prefetch_L1(x) __insn_prefetch_L1((void *)(x))
#endif
#else /* __ASSEMBLY__ */
/* Do some slow action (e.g. read a slow SPR). */

300
arch/tile/include/hv/drv_xgbe_impl.h Normal file
View file

@ -0,0 +1,300 @@
/*
* Copyright 2010 Tilera Corporation. All Rights Reserved.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation, version 2.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
* NON INFRINGEMENT. See the GNU General Public License for
* more details.
*/
/**
* @file drivers/xgbe/impl.h
* Implementation details for the NetIO library.
*/
#ifndef __DRV_XGBE_IMPL_H__
#define __DRV_XGBE_IMPL_H__
#include <hv/netio_errors.h>
#include <hv/netio_intf.h>
#include <hv/drv_xgbe_intf.h>
/** How many groups we have (log2). */
#define LOG2_NUM_GROUPS (12)
/** How many groups we have. */
#define NUM_GROUPS (1 << LOG2_NUM_GROUPS)
/** Number of output requests we'll buffer per tile. */
#define EPP_REQS_PER_TILE (32)
/** Words used in an eDMA command without checksum acceleration. */
#define EDMA_WDS_NO_CSUM 8
/** Words used in an eDMA command with checksum acceleration. */
#define EDMA_WDS_CSUM 10
/** Total available words in the eDMA command FIFO. */
#define EDMA_WDS_TOTAL 128
/*
* FIXME: These definitions are internal and should have underscores!
* NOTE: The actual numeric values here are intentional and allow us to
* optimize the concept "if small ... else if large ... else ...", by
* checking for the low bit being set, and then for non-zero.
* These are used as array indices, so they must have the values (0, 1, 2)
* in some order.
*/
#define SIZE_SMALL (1) /**< Small packet queue. */
#define SIZE_LARGE (2) /**< Large packet queue. */
#define SIZE_JUMBO (0) /**< Jumbo packet queue. */
/** The number of "SIZE_xxx" values. */
#define NETIO_NUM_SIZES 3
/*
* Default numbers of packets for IPP drivers. These values are chosen
* such that CIPP1 will not overflow its L2 cache.
*/
/** The default number of small packets. */
#define NETIO_DEFAULT_SMALL_PACKETS 2750
/** The default number of large packets. */
#define NETIO_DEFAULT_LARGE_PACKETS 2500
/** The default number of jumbo packets. */
#define NETIO_DEFAULT_JUMBO_PACKETS 250
/** Log2 of the size of a memory arena. */
#define NETIO_ARENA_SHIFT 24 /* 16 MB */
/** Size of a memory arena. */
#define NETIO_ARENA_SIZE (1 << NETIO_ARENA_SHIFT)
/** A queue of packets.
*
* This structure partially defines a queue of packets waiting to be
* processed. The queue as a whole is written to by an interrupt handler and
* read by non-interrupt code; this data structure is what's touched by the
* interrupt handler. The other part of the queue state, the read offset, is
* kept in user space, not in hypervisor space, so it is in a separate data
* structure.
*
* The read offset (__packet_receive_read in the user part of the queue
* structure) points to the next packet to be read. When the read offset is
* equal to the write offset, the queue is empty; therefore the queue must
* contain one more slot than the required maximum queue size.
*
* Here's an example of all 3 state variables and what they mean. All
* pointers move left to right.
*
* @code
* I I V V V V I I I I
* 0 1 2 3 4 5 6 7 8 9 10
* ^ ^ ^ ^
* | | |
* | | __last_packet_plus_one
* | __buffer_write
* __packet_receive_read
* @endcode
*
* This queue has 10 slots, and thus can hold 9 packets (_last_packet_plus_one
* = 10). The read pointer is at 2, and the write pointer is at 6; thus,
* there are valid, unread packets in slots 2, 3, 4, and 5. The remaining
* slots are invalid (do not contain a packet).
*/
typedef struct {
/** Byte offset of the next notify packet to be written: zero for the first
* packet on the queue, sizeof (netio_pkt_t) for the second packet on the
* queue, etc. */
volatile uint32_t __packet_write;
/** Offset of the packet after the last valid packet (i.e., when any
* pointer is incremented to this value, it wraps back to zero). */
uint32_t __last_packet_plus_one;
}
__netio_packet_queue_t;
/** A queue of buffers.
*
* This structure partially defines a queue of empty buffers which have been
* obtained via requests to the IPP. (The elements of the queue are packet
* handles, which are transformed into a full netio_pkt_t when the buffer is
* retrieved.) The queue as a whole is written to by an interrupt handler and
* read by non-interrupt code; this data structure is what's touched by the
* interrupt handler. The other parts of the queue state, the read offset and
* requested write offset, are kept in user space, not in hypervisor space, so
* they are in a separate data structure.
*
* The read offset (__buffer_read in the user part of the queue structure)
* points to the next buffer to be read. When the read offset is equal to the
* write offset, the queue is empty; therefore the queue must contain one more
* slot than the required maximum queue size.
*
* The requested write offset (__buffer_requested_write in the user part of
* the queue structure) points to the slot which will hold the next buffer we
* request from the IPP, once we get around to sending such a request. When
* the requested write offset is equal to the write offset, no requests for
* new buffers are outstanding; when the requested write offset is one greater
* than the read offset, no more requests may be sent.
*
* Note that, unlike the packet_queue, the buffer_queue places incoming
* buffers at decreasing addresses. This makes the check for "is it time to
* wrap the buffer pointer" cheaper in the assembly code which receives new
* buffers, and means that the value which defines the queue size,
* __last_buffer, is different than in the packet queue. Also, the offset
* used in the packet_queue is already scaled by the size of a packet; here we
* use unscaled slot indices for the offsets. (These differences are
* historical, and in the future it's possible that the packet_queue will look
* more like this queue.)
*
* @code
* Here's an example of all 4 state variables and what they mean. Remember:
* all pointers move right to left.
*
* V V V I I R R V V V
* 0 1 2 3 4 5 6 7 8 9
* ^ ^ ^ ^
* | | | |
* | | | __last_buffer
* | | __buffer_write
* | __buffer_requested_write
* __buffer_read
* @endcode
*
* This queue has 10 slots, and thus can hold 9 buffers (_last_buffer = 9).
* The read pointer is at 2, and the write pointer is at 6; thus, there are
* valid, unread buffers in slots 2, 1, 0, 9, 8, and 7. The requested write
* pointer is at 4; thus, requests have been made to the IPP for buffers which
* will be placed in slots 6 and 5 when they arrive. Finally, the remaining
* slots are invalid (do not contain a buffer).
*/
typedef struct
{
/** Ordinal number of the next buffer to be written: 0 for the first slot in
* the queue, 1 for the second slot in the queue, etc. */
volatile uint32_t __buffer_write;
/** Ordinal number of the last buffer (i.e., when any pointer is decremented
* below zero, it is reloaded with this value). */
uint32_t __last_buffer;
}
__netio_buffer_queue_t;
/**
* An object for providing Ethernet packets to a process.
*/
typedef struct __netio_queue_impl_t
{
/** The queue of packets waiting to be received. */
__netio_packet_queue_t __packet_receive_queue;
/** The intr bit mask that IDs this device. */
unsigned int __intr_id;
/** Offset to queues of empty buffers, one per size. */
uint32_t __buffer_queue[NETIO_NUM_SIZES];
/** The address of the first EPP tile, or -1 if no EPP. */
/* ISSUE: Actually this is always "0" or "~0". */
uint32_t __epp_location;
/** The queue ID that this queue represents. */
unsigned int __queue_id;
/** Number of acknowledgements received. */
volatile uint32_t __acks_received;
/** Last completion number received for packet_sendv. */
volatile uint32_t __last_completion_rcv;
/** Number of packets allowed to be outstanding. */
uint32_t __max_outstanding;
/** First VA available for packets. */
void* __va_0;
/** First VA in second range available for packets. */
void* __va_1;
/** Padding to align the "__packets" field to the size of a netio_pkt_t. */
uint32_t __padding[3];
/** The packets themselves. */
netio_pkt_t __packets[0];
}
netio_queue_impl_t;
/**
* An object for managing the user end of a NetIO queue.
*/
typedef struct __netio_queue_user_impl_t
{
/** The next incoming packet to be read. */
uint32_t __packet_receive_read;
/** The next empty buffers to be read, one index per size. */
uint8_t __buffer_read[NETIO_NUM_SIZES];
/** Where the empty buffer we next request from the IPP will go, one index
* per size. */
uint8_t __buffer_requested_write[NETIO_NUM_SIZES];
/** PCIe interface flag. */
uint8_t __pcie;
/** Number of packets left to be received before we send a credit update. */
uint32_t __receive_credit_remaining;
/** Value placed in __receive_credit_remaining when it reaches zero. */
uint32_t __receive_credit_interval;
/** First fast I/O routine index. */
uint32_t __fastio_index;
/** Number of acknowledgements expected. */
uint32_t __acks_outstanding;
/** Last completion number requested. */
uint32_t __last_completion_req;
/** File descriptor for driver. */
int __fd;
}
netio_queue_user_impl_t;
#define NETIO_GROUP_CHUNK_SIZE 64 /**< Max # groups in one IPP request */
#define NETIO_BUCKET_CHUNK_SIZE 64 /**< Max # buckets in one IPP request */
/** Internal structure used to convey packet send information to the
* hypervisor. FIXME: Actually, it's not used for that anymore, but
* netio_packet_send() still uses it internally.
*/
typedef struct
{
uint16_t flags; /**< Packet flags (__NETIO_SEND_FLG_xxx) */
uint16_t transfer_size; /**< Size of packet */
uint32_t va; /**< VA of start of packet */
__netio_pkt_handle_t handle; /**< Packet handle */
uint32_t csum0; /**< First checksum word */
uint32_t csum1; /**< Second checksum word */
}
__netio_send_cmd_t;
/** Flags used in two contexts:
* - As the "flags" member in the __netio_send_cmd_t, above; used only
* for netio_pkt_send_{prepare,commit}.
* - As part of the flags passed to the various send packet fast I/O calls.
*/
/** Need acknowledgement on this packet. Note that some code in the
* normal send_pkt fast I/O handler assumes that this is equal to 1. */
#define __NETIO_SEND_FLG_ACK 0x1
/** Do checksum on this packet. (Only used with the __netio_send_cmd_t;
* normal packet sends use a special fast I/O index to denote checksumming,
* and multi-segment sends test the checksum descriptor.) */
#define __NETIO_SEND_FLG_CSUM 0x2
/** Get a completion on this packet. Only used with multi-segment sends. */
#define __NETIO_SEND_FLG_COMPLETION 0x4
/** Position of the number-of-extra-segments value in the flags word.
Only used with multi-segment sends. */
#define __NETIO_SEND_FLG_XSEG_SHIFT 3
/** Width of the number-of-extra-segments value in the flags word. */
#define __NETIO_SEND_FLG_XSEG_WIDTH 2
#endif /* __DRV_XGBE_IMPL_H__ */

615
arch/tile/include/hv/drv_xgbe_intf.h Normal file
View file

@ -0,0 +1,615 @@
/*
* Copyright 2010 Tilera Corporation. All Rights Reserved.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation, version 2.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
* NON INFRINGEMENT. See the GNU General Public License for
* more details.
*/
/**
* @file drv_xgbe_intf.h
* Interface to the hypervisor XGBE driver.
*/
#ifndef __DRV_XGBE_INTF_H__
#define __DRV_XGBE_INTF_H__
/**
* An object for forwarding VAs and PAs to the hypervisor.
* @ingroup types
*
* This allows the supervisor to specify a number of areas of memory to
* store packet buffers.
*/
typedef struct
{
/** The physical address of the memory. */
HV_PhysAddr pa;
/** Page table entry for the memory. This is only used to derive the
* memory's caching mode; the PA bits are ignored. */
HV_PTE pte;
/** The virtual address of the memory. */
HV_VirtAddr va;
/** Size (in bytes) of the memory area. */
int size;
}
netio_ipp_address_t;
/** The various pread/pwrite offsets into the hypervisor-level driver.
* @ingroup types
*/
typedef enum
{
/** Inform the Linux driver of the address of the NetIO arena memory.
* This offset is actually only used to convey information from netio
* to the Linux driver; it never makes it from there to the hypervisor.
* Write-only; takes a uint32_t specifying the VA address. */
NETIO_FIXED_ADDR = 0x5000000000000000ULL,
/** Inform the Linux driver of the size of the NetIO arena memory.
* This offset is actually only used to convey information from netio
* to the Linux driver; it never makes it from there to the hypervisor.
* Write-only; takes a uint32_t specifying the VA size. */
NETIO_FIXED_SIZE = 0x5100000000000000ULL,
/** Register current tile with IPP. Write then read: write, takes a
* netio_input_config_t, read returns a pointer to a netio_queue_impl_t. */
NETIO_IPP_INPUT_REGISTER_OFF = 0x6000000000000000ULL,
/** Unregister current tile from IPP. Write-only, takes a dummy argument. */
NETIO_IPP_INPUT_UNREGISTER_OFF = 0x6100000000000000ULL,
/** Start packets flowing. Write-only, takes a dummy argument. */
NETIO_IPP_INPUT_INIT_OFF = 0x6200000000000000ULL,
/** Stop packets flowing. Write-only, takes a dummy argument. */
NETIO_IPP_INPUT_UNINIT_OFF = 0x6300000000000000ULL,
/** Configure group (typically we group on VLAN). Write-only: takes an
* array of netio_group_t's, low 24 bits of the offset is the base group
* number times the size of a netio_group_t. */
NETIO_IPP_INPUT_GROUP_CFG_OFF = 0x6400000000000000ULL,
/** Configure bucket. Write-only: takes an array of netio_bucket_t's, low
* 24 bits of the offset is the base bucket number times the size of a
* netio_bucket_t. */
NETIO_IPP_INPUT_BUCKET_CFG_OFF = 0x6500000000000000ULL,
/** Get/set a parameter. Read or write: read or write data is the parameter
* value, low 32 bits of the offset is a __netio_getset_offset_t. */
NETIO_IPP_PARAM_OFF = 0x6600000000000000ULL,
/** Get fast I/O index. Read-only; returns a 4-byte base index value. */
NETIO_IPP_GET_FASTIO_OFF = 0x6700000000000000ULL,
/** Configure hijack IP address. Packets with this IPv4 dest address
* go to bucket NETIO_NUM_BUCKETS - 1. Write-only: takes an IP address
* in some standard form. FIXME: Define the form! */
NETIO_IPP_INPUT_HIJACK_CFG_OFF = 0x6800000000000000ULL,
/**
* Offsets beyond this point are reserved for the supervisor (although that
* enforcement must be done by the supervisor driver itself).
*/
NETIO_IPP_USER_MAX_OFF = 0x6FFFFFFFFFFFFFFFULL,
/** Register I/O memory. Write-only, takes a netio_ipp_address_t. */
NETIO_IPP_IOMEM_REGISTER_OFF = 0x7000000000000000ULL,
/** Unregister I/O memory. Write-only, takes a netio_ipp_address_t. */
NETIO_IPP_IOMEM_UNREGISTER_OFF = 0x7100000000000000ULL,
/* Offsets greater than 0x7FFFFFFF can't be used directly from Linux
* userspace code due to limitations in the pread/pwrite syscalls. */
/** Drain LIPP buffers. */
NETIO_IPP_DRAIN_OFF = 0xFA00000000000000ULL,
/** Supply a netio_ipp_address_t to be used as shared memory for the
* LEPP command queue. */
NETIO_EPP_SHM_OFF = 0xFB00000000000000ULL,
/* 0xFC... is currently unused. */
/** Stop IPP/EPP tiles. Write-only, takes a dummy argument. */
NETIO_IPP_STOP_SHIM_OFF = 0xFD00000000000000ULL,
/** Start IPP/EPP tiles. Write-only, takes a dummy argument. */
NETIO_IPP_START_SHIM_OFF = 0xFE00000000000000ULL,
/** Supply packet arena. Write-only, takes an array of
* netio_ipp_address_t values. */
NETIO_IPP_ADDRESS_OFF = 0xFF00000000000000ULL,
} netio_hv_offset_t;
/** Extract the base offset from an offset */
#define NETIO_BASE_OFFSET(off) ((off) & 0xFF00000000000000ULL)
/** Extract the local offset from an offset */
#define NETIO_LOCAL_OFFSET(off) ((off) & 0x00FFFFFFFFFFFFFFULL)
/**
* Get/set offset.
*/
typedef union
{
struct
{
uint64_t addr:48; /**< Class-specific address */
unsigned int class:8; /**< Class (e.g., NETIO_PARAM) */
unsigned int opcode:8; /**< High 8 bits of NETIO_IPP_PARAM_OFF */
}
bits; /**< Bitfields */
uint64_t word; /**< Aggregated value to use as the offset */
}
__netio_getset_offset_t;
/**
* Fast I/O index offsets (must be contiguous).
*/
typedef enum
{
NETIO_FASTIO_ALLOCATE = 0, /**< Get empty packet buffer */
NETIO_FASTIO_FREE_BUFFER = 1, /**< Give buffer back to IPP */
NETIO_FASTIO_RETURN_CREDITS = 2, /**< Give credits to IPP */
NETIO_FASTIO_SEND_PKT_NOCK = 3, /**< Send a packet, no checksum */
NETIO_FASTIO_SEND_PKT_CK = 4, /**< Send a packet, with checksum */
NETIO_FASTIO_SEND_PKT_VEC = 5, /**< Send a vector of packets */
NETIO_FASTIO_SENDV_PKT = 6, /**< Sendv one packet */
NETIO_FASTIO_NUM_INDEX = 7, /**< Total number of fast I/O indices */
} netio_fastio_index_t;
/** 3-word return type for Fast I/O call. */
typedef struct
{
int err; /**< Error code. */
uint32_t val0; /**< Value. Meaning depends upon the specific call. */
uint32_t val1; /**< Value. Meaning depends upon the specific call. */
} netio_fastio_rv3_t;
/** 0-argument fast I/O call */
int __netio_fastio0(uint32_t fastio_index);
/** 1-argument fast I/O call */
int __netio_fastio1(uint32_t fastio_index, uint32_t arg0);
/** 3-argument fast I/O call, 2-word return value */
netio_fastio_rv3_t __netio_fastio3_rv3(uint32_t fastio_index, uint32_t arg0,
uint32_t arg1, uint32_t arg2);
/** 4-argument fast I/O call */
int __netio_fastio4(uint32_t fastio_index, uint32_t arg0, uint32_t arg1,
uint32_t arg2, uint32_t arg3);
/** 6-argument fast I/O call */
int __netio_fastio6(uint32_t fastio_index, uint32_t arg0, uint32_t arg1,
uint32_t arg2, uint32_t arg3, uint32_t arg4, uint32_t arg5);
/** 9-argument fast I/O call */
int __netio_fastio9(uint32_t fastio_index, uint32_t arg0, uint32_t arg1,
uint32_t arg2, uint32_t arg3, uint32_t arg4, uint32_t arg5,
uint32_t arg6, uint32_t arg7, uint32_t arg8);
/** Allocate an empty packet.
* @param fastio_index Fast I/O index.
* @param size Size of the packet to allocate.
*/
#define __netio_fastio_allocate(fastio_index, size) \
__netio_fastio1((fastio_index) + NETIO_FASTIO_ALLOCATE, size)
/** Free a buffer.
* @param fastio_index Fast I/O index.
* @param handle Handle for the packet to free.
*/
#define __netio_fastio_free_buffer(fastio_index, handle) \
__netio_fastio1((fastio_index) + NETIO_FASTIO_FREE_BUFFER, handle)
/** Increment our receive credits.
* @param fastio_index Fast I/O index.
* @param credits Number of credits to add.
*/
#define __netio_fastio_return_credits(fastio_index, credits) \
__netio_fastio1((fastio_index) + NETIO_FASTIO_RETURN_CREDITS, credits)
/** Send packet, no checksum.
* @param fastio_index Fast I/O index.
* @param ackflag Nonzero if we want an ack.
* @param size Size of the packet.
* @param va Virtual address of start of packet.
* @param handle Packet handle.
*/
#define __netio_fastio_send_pkt_nock(fastio_index, ackflag, size, va, handle) \
__netio_fastio4((fastio_index) + NETIO_FASTIO_SEND_PKT_NOCK, ackflag, \
size, va, handle)
/** Send packet, calculate checksum.
* @param fastio_index Fast I/O index.
* @param ackflag Nonzero if we want an ack.
* @param size Size of the packet.
* @param va Virtual address of start of packet.
* @param handle Packet handle.
* @param csum0 Shim checksum header.
* @param csum1 Checksum seed.
*/
#define __netio_fastio_send_pkt_ck(fastio_index, ackflag, size, va, handle, \
csum0, csum1) \
__netio_fastio6((fastio_index) + NETIO_FASTIO_SEND_PKT_CK, ackflag, \
size, va, handle, csum0, csum1)
/** Format for the "csum0" argument to the __netio_fastio_send routines
* and LEPP. Note that this is currently exactly identical to the
* ShimProtocolOffloadHeader.
*/
typedef union
{
struct
{
unsigned int start_byte:7; /**< The first byte to be checksummed */
unsigned int count:14; /**< Number of bytes to be checksummed. */
unsigned int destination_byte:7; /**< The byte to write the checksum to. */
unsigned int reserved:4; /**< Reserved. */
} bits; /**< Decomposed method of access. */
unsigned int word; /**< To send out the IDN. */
} __netio_checksum_header_t;
/** Sendv packet with 1 or 2 segments.
* @param fastio_index Fast I/O index.
* @param flags Ack/csum/notify flags in low 3 bits; number of segments minus
* 1 in next 2 bits; expected checksum in high 16 bits.
* @param confno Confirmation number to request, if notify flag set.
* @param csum0 Checksum descriptor; if zero, no checksum.
* @param va_F Virtual address of first segment.
* @param va_L Virtual address of last segment, if 2 segments.
* @param len_F_L Length of first segment in low 16 bits; length of last
* segment, if 2 segments, in high 16 bits.
*/
#define __netio_fastio_sendv_pkt_1_2(fastio_index, flags, confno, csum0, \
va_F, va_L, len_F_L) \
__netio_fastio6((fastio_index) + NETIO_FASTIO_SENDV_PKT, flags, confno, \
csum0, va_F, va_L, len_F_L)
/** Send packet on PCIe interface.
* @param fastio_index Fast I/O index.
* @param flags Ack/csum/notify flags in low 3 bits.
* @param confno Confirmation number to request, if notify flag set.
* @param csum0 Checksum descriptor; Hard wired 0, not needed for PCIe.
* @param va_F Virtual address of the packet buffer.
* @param va_L Virtual address of last segment, if 2 segments. Hard wired 0.
* @param len_F_L Length of the packet buffer in low 16 bits.
*/
#define __netio_fastio_send_pcie_pkt(fastio_index, flags, confno, csum0, \
va_F, va_L, len_F_L) \
__netio_fastio6((fastio_index) + PCIE_FASTIO_SENDV_PKT, flags, confno, \
csum0, va_F, va_L, len_F_L)
/** Sendv packet with 3 or 4 segments.
* @param fastio_index Fast I/O index.
* @param flags Ack/csum/notify flags in low 3 bits; number of segments minus
* 1 in next 2 bits; expected checksum in high 16 bits.
* @param confno Confirmation number to request, if notify flag set.
* @param csum0 Checksum descriptor; if zero, no checksum.
* @param va_F Virtual address of first segment.
* @param va_L Virtual address of last segment (third segment if 3 segments,
* fourth segment if 4 segments).
* @param len_F_L Length of first segment in low 16 bits; length of last
* segment in high 16 bits.
* @param va_M0 Virtual address of "middle 0" segment; this segment is sent
* second when there are three segments, and third if there are four.
* @param va_M1 Virtual address of "middle 1" segment; this segment is sent
* second when there are four segments.
* @param len_M0_M1 Length of middle 0 segment in low 16 bits; length of middle
* 1 segment, if 4 segments, in high 16 bits.
*/
#define __netio_fastio_sendv_pkt_3_4(fastio_index, flags, confno, csum0, va_F, \
va_L, len_F_L, va_M0, va_M1, len_M0_M1) \
__netio_fastio9((fastio_index) + NETIO_FASTIO_SENDV_PKT, flags, confno, \
csum0, va_F, va_L, len_F_L, va_M0, va_M1, len_M0_M1)
/** Send vector of packets.
* @param fastio_index Fast I/O index.
* @param seqno Number of packets transmitted so far on this interface;
* used to decide which packets should be acknowledged.
* @param nentries Number of entries in vector.
* @param va Virtual address of start of vector entry array.
* @return 3-word netio_fastio_rv3_t structure. The structure's err member
* is an error code, or zero if no error. The val0 member is the
* updated value of seqno; it has been incremented by 1 for each
* packet sent. That increment may be less than nentries if an
* error occured, or if some of the entries in the vector contain
* handles equal to NETIO_PKT_HANDLE_NONE. The val1 member is the
* updated value of nentries; it has been decremented by 1 for each
* vector entry processed. Again, that decrement may be less than
* nentries (leaving the returned value positive) if an error
* occurred.
*/
#define __netio_fastio_send_pkt_vec(fastio_index, seqno, nentries, va) \
__netio_fastio3_rv3((fastio_index) + NETIO_FASTIO_SEND_PKT_VEC, seqno, \
nentries, va)
/** An egress DMA command for LEPP. */
typedef struct
{
/** Is this a TSO transfer?
*
* NOTE: This field is always 0, to distinguish it from
* lepp_tso_cmd_t. It must come first!
*/
uint8_t tso : 1;
/** Unused padding bits. */
uint8_t _unused : 3;
/** Should this packet be sent directly from caches instead of DRAM,
* using hash-for-home to locate the packet data?
*/
uint8_t hash_for_home : 1;
/** Should we compute a checksum? */
uint8_t compute_checksum : 1;
/** Is this the final buffer for this packet?
*
* A single packet can be split over several input buffers (a "gather"
* operation). This flag indicates that this is the last buffer
* in a packet.
*/
uint8_t end_of_packet : 1;
/** Should LEPP advance 'comp_busy' when this DMA is fully finished? */
uint8_t send_completion : 1;
/** High bits of Client Physical Address of the start of the buffer
* to be egressed.
*
* NOTE: Only 6 bits are actually needed here, as CPAs are
* currently 38 bits. So two bits could be scavenged from this.
*/
uint8_t cpa_hi;
/** The number of bytes to be egressed. */
uint16_t length;
/** Low 32 bits of Client Physical Address of the start of the buffer
* to be egressed.
*/
uint32_t cpa_lo;
/** Checksum information (only used if 'compute_checksum'). */
__netio_checksum_header_t checksum_data;
} lepp_cmd_t;
/** A chunk of physical memory for a TSO egress. */
typedef struct
{
/** The low bits of the CPA. */
uint32_t cpa_lo;
/** The high bits of the CPA. */
uint16_t cpa_hi : 15;
/** Should this packet be sent directly from caches instead of DRAM,
* using hash-for-home to locate the packet data?
*/
uint16_t hash_for_home : 1;
/** The length in bytes. */
uint16_t length;
} lepp_frag_t;
/** An LEPP command that handles TSO. */
typedef struct
{
/** Is this a TSO transfer?
*
* NOTE: This field is always 1, to distinguish it from
* lepp_cmd_t. It must come first!
*/
uint8_t tso : 1;
/** Unused padding bits. */
uint8_t _unused : 7;
/** Size of the header[] array in bytes. It must be in the range
* [40, 127], which are the smallest header for a TCP packet over
* Ethernet and the maximum possible prepend size supported by
* hardware, respectively. Note that the array storage must be
* padded out to a multiple of four bytes so that the following
* LEPP command is aligned properly.
*/
uint8_t header_size;
/** Byte offset of the IP header in header[]. */
uint8_t ip_offset;
/** Byte offset of the TCP header in header[]. */
uint8_t tcp_offset;
/** The number of bytes to use for the payload of each packet,
* except of course the last one, which may not have enough bytes.
* This means that each Ethernet packet except the last will have a
* size of header_size + payload_size.
*/
uint16_t payload_size;
/** The length of the 'frags' array that follows this struct. */
uint16_t num_frags;
/** The actual frags. */
lepp_frag_t frags[0 /* Variable-sized; num_frags entries. */];
/*
* The packet header template logically follows frags[],
* but you can't declare that in C.
*
* uint32_t header[header_size_in_words_rounded_up];
*/
} lepp_tso_cmd_t;
/** An LEPP completion ring entry. */
typedef void* lepp_comp_t;
/** Maximum number of frags for one TSO command. This is adapted from
* linux's "MAX_SKB_FRAGS", and presumably over-estimates by one, for
* our page size of exactly 65536. We add one for a "body" fragment.
*/
#define LEPP_MAX_FRAGS (65536 / HV_PAGE_SIZE_SMALL + 2 + 1)
/** Total number of bytes needed for an lepp_tso_cmd_t. */
#define LEPP_TSO_CMD_SIZE(num_frags, header_size) \
(sizeof(lepp_tso_cmd_t) + \
(num_frags) * sizeof(lepp_frag_t) + \
(((header_size) + 3) & -4))
/** The size of the lepp "cmd" queue. */
#define LEPP_CMD_QUEUE_BYTES \
(((CHIP_L2_CACHE_SIZE() - 2 * CHIP_L2_LINE_SIZE()) / \
(sizeof(lepp_cmd_t) + sizeof(lepp_comp_t))) * sizeof(lepp_cmd_t))
/** The largest possible command that can go in lepp_queue_t::cmds[]. */
#define LEPP_MAX_CMD_SIZE LEPP_TSO_CMD_SIZE(LEPP_MAX_FRAGS, 128)
/** The largest possible value of lepp_queue_t::cmd_{head, tail} (inclusive).
*/
#define LEPP_CMD_LIMIT \
(LEPP_CMD_QUEUE_BYTES - LEPP_MAX_CMD_SIZE)
/** The maximum number of completions in an LEPP queue. */
#define LEPP_COMP_QUEUE_SIZE \
((LEPP_CMD_LIMIT + sizeof(lepp_cmd_t) - 1) / sizeof(lepp_cmd_t))
/** Increment an index modulo the queue size. */
#define LEPP_QINC(var) \
(var = __insn_mnz(var - (LEPP_COMP_QUEUE_SIZE - 1), var + 1))
/** A queue used to convey egress commands from the client to LEPP. */
typedef struct
{
/** Index of first completion not yet processed by user code.
* If this is equal to comp_busy, there are no such completions.
*
* NOTE: This is only read/written by the user.
*/
unsigned int comp_head;
/** Index of first completion record not yet completed.
* If this is equal to comp_tail, there are no such completions.
* This index gets advanced (modulo LEPP_QUEUE_SIZE) whenever
* a command with the 'completion' bit set is finished.
*
* NOTE: This is only written by LEPP, only read by the user.
*/
volatile unsigned int comp_busy;
/** Index of the first empty slot in the completion ring.
* Entries from this up to but not including comp_head (in ring order)
* can be filled in with completion data.
*
* NOTE: This is only read/written by the user.
*/
unsigned int comp_tail;
/** Byte index of first command enqueued for LEPP but not yet processed.
*
* This is always divisible by sizeof(void*) and always <= LEPP_CMD_LIMIT.
*
* NOTE: LEPP advances this counter as soon as it no longer needs
* the cmds[] storage for this entry, but the transfer is not actually
* complete (i.e. the buffer pointed to by the command is no longer
* needed) until comp_busy advances.
*
* If this is equal to cmd_tail, the ring is empty.
*
* NOTE: This is only written by LEPP, only read by the user.
*/
volatile unsigned int cmd_head;
/** Byte index of first empty slot in the command ring. This field can
* be incremented up to but not equal to cmd_head (because that would
* mean the ring is empty).
*
* This is always divisible by sizeof(void*) and always <= LEPP_CMD_LIMIT.
*
* NOTE: This is read/written by the user, only read by LEPP.
*/
volatile unsigned int cmd_tail;
/** A ring of variable-sized egress DMA commands.
*
* NOTE: Only written by the user, only read by LEPP.
*/
char cmds[LEPP_CMD_QUEUE_BYTES]
__attribute__((aligned(CHIP_L2_LINE_SIZE())));
/** A ring of user completion data.
* NOTE: Only read/written by the user.
*/
lepp_comp_t comps[LEPP_COMP_QUEUE_SIZE]
__attribute__((aligned(CHIP_L2_LINE_SIZE())));
} lepp_queue_t;
/** An internal helper function for determining the number of entries
* available in a ring buffer, given that there is one sentinel.
*/
static inline unsigned int
_lepp_num_free_slots(unsigned int head, unsigned int tail)
{
/*
* One entry is reserved for use as a sentinel, to distinguish
* "empty" from "full". So we compute
* (head - tail - 1) % LEPP_QUEUE_SIZE, but without using a slow % operation.
*/
return (head - tail - 1) + ((head <= tail) ? LEPP_COMP_QUEUE_SIZE : 0);
}
/** Returns how many new comp entries can be enqueued. */
static inline unsigned int
lepp_num_free_comp_slots(const lepp_queue_t* q)
{
return _lepp_num_free_slots(q->comp_head, q->comp_tail);
}
static inline int
lepp_qsub(int v1, int v2)
{
int delta = v1 - v2;
return delta + ((delta >> 31) & LEPP_COMP_QUEUE_SIZE);
}
/** FIXME: Check this from linux, via a new "pwrite()" call. */
#define LIPP_VERSION 1
/** We use exactly two bytes of alignment padding. */
#define LIPP_PACKET_PADDING 2
/** The minimum size of a "small" buffer (including the padding). */
#define LIPP_SMALL_PACKET_SIZE 128
/*
* NOTE: The following two values should total to less than around
* 13582, to keep the total size used for "lipp_state_t" below 64K.
*/
/** The maximum number of "small" buffers.
* This is enough for 53 network cpus with 128 credits. Note that
* if these are exhausted, we will fall back to using large buffers.
*/
#define LIPP_SMALL_BUFFERS 6785
/** The maximum number of "large" buffers.
* This is enough for 53 network cpus with 128 credits.
*/
#define LIPP_LARGE_BUFFERS 6785
#endif /* __DRV_XGBE_INTF_H__ */

122
arch/tile/include/hv/netio_errors.h Normal file
View file

@ -0,0 +1,122 @@
/*
* Copyright 2010 Tilera Corporation. All Rights Reserved.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation, version 2.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
* NON INFRINGEMENT. See the GNU General Public License for
* more details.
*/
/**
* Error codes returned from NetIO routines.
*/
#ifndef __NETIO_ERRORS_H__
#define __NETIO_ERRORS_H__
/**
* @addtogroup error
*
* @brief The error codes returned by NetIO functions.
*
* NetIO functions return 0 (defined as ::NETIO_NO_ERROR) on success, and
* a negative value if an error occurs.
*
* In cases where a NetIO function failed due to a error reported by
* system libraries, the error code will be the negation of the
* system errno at the time of failure. The @ref netio_strerror()
* function will deliver error strings for both NetIO and system error
* codes.
*
* @{
*/
/** The set of all NetIO errors. */
typedef enum
{
/** Operation successfully completed. */
NETIO_NO_ERROR = 0,
/** A packet was successfully retrieved from an input queue. */
NETIO_PKT = 0,
/** Largest NetIO error number. */
NETIO_ERR_MAX = -701,
/** The tile is not registered with the IPP. */
NETIO_NOT_REGISTERED = -701,
/** No packet was available to retrieve from the input queue. */
NETIO_NOPKT = -702,
/** The requested function is not implemented. */
NETIO_NOT_IMPLEMENTED = -703,
/** On a registration operation, the target queue already has the maximum
* number of tiles registered for it, and no more may be added. On a
* packet send operation, the output queue is full and nothing more can
* be queued until some of the queued packets are actually transmitted. */
NETIO_QUEUE_FULL = -704,
/** The calling process or thread is not bound to exactly one CPU. */
NETIO_BAD_AFFINITY = -705,
/** Cannot allocate memory on requested controllers. */
NETIO_CANNOT_HOME = -706,
/** On a registration operation, the IPP specified is not configured
* to support the options requested; for instance, the application
* wants a specific type of tagged headers which the configured IPP
* doesn't support. Or, the supplied configuration information is
* not self-consistent, or is out of range; for instance, specifying
* both NETIO_RECV and NETIO_NO_RECV, or asking for more than
* NETIO_MAX_SEND_BUFFERS to be preallocated. On a VLAN or bucket
* configure operation, the number of items, or the base item, was
* out of range.
*/
NETIO_BAD_CONFIG = -707,
/** Too many tiles have registered to transmit packets. */
NETIO_TOOMANY_XMIT = -708,
/** Packet transmission was attempted on a queue which was registered
with transmit disabled. */
NETIO_UNREG_XMIT = -709,
/** This tile is already registered with the IPP. */
NETIO_ALREADY_REGISTERED = -710,
/** The Ethernet link is down. The application should try again later. */
NETIO_LINK_DOWN = -711,
/** An invalid memory buffer has been specified. This may be an unmapped
* virtual address, or one which does not meet alignment requirements.
* For netio_input_register(), this error may be returned when multiple
* processes specify different memory regions to be used for NetIO
* buffers. That can happen if these processes specify explicit memory
* regions with the ::NETIO_FIXED_BUFFER_VA flag, or if tmc_cmem_init()
* has not been called by a common ancestor of the processes.
*/
NETIO_FAULT = -712,
/** Cannot combine user-managed shared memory and cache coherence. */
NETIO_BAD_CACHE_CONFIG = -713,
/** Smallest NetIO error number. */
NETIO_ERR_MIN = -713,
#ifndef __DOXYGEN__
/** Used internally to mean that no response is needed; never returned to
* an application. */
NETIO_NO_RESPONSE = 1
#endif
} netio_error_t;
/** @} */
#endif /* __NETIO_ERRORS_H__ */

2975
arch/tile/include/hv/netio_intf.h Normal file
View file

File diff suppressed because it is too large Load diff

1
arch/tile/kernel/Makefile
View file

@ -15,3 +15,4 @@ obj-$(CONFIG_SMP) += smpboot.o smp.o tlb.o
obj-$(CONFIG_MODULES) += module.o
obj-$(CONFIG_EARLY_PRINTK) += early_printk.o
obj-$(CONFIG_KEXEC) += machine_kexec.o relocate_kernel.o
obj-$(CONFIG_PCI) += pci.o

621
arch/tile/kernel/pci.c Normal file
View file

@ -0,0 +1,621 @@
/*
* Copyright 2010 Tilera Corporation. All Rights Reserved.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation, version 2.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, GOOD TITLE or
* NON INFRINGEMENT. See the GNU General Public License for
* more details.
*/
#include <linux/kernel.h>
#include <linux/pci.h>
#include <linux/delay.h>
#include <linux/string.h>
#include <linux/init.h>
#include <linux/capability.h>
#include <linux/sched.h>
#include <linux/errno.h>
#include <linux/bootmem.h>
#include <linux/irq.h>
#include <linux/io.h>
#include <linux/uaccess.h>
#include <asm/processor.h>
#include <asm/sections.h>
#include <asm/byteorder.h>
#include <asm/hv_driver.h>
#include <hv/drv_pcie_rc_intf.h>
/*
* Initialization flow and process
* -------------------------------
*
* This files containes the routines to search for PCI buses,
* enumerate the buses, and configure any attached devices.
*
* There are two entry points here:
* 1) tile_pci_init
* This sets up the pci_controller structs, and opens the
* FDs to the hypervisor. This is called from setup_arch() early
* in the boot process.
* 2) pcibios_init
* This probes the PCI bus(es) for any attached hardware. It's
* called by subsys_initcall. All of the real work is done by the
* generic Linux PCI layer.
*
*/
/*
* This flag tells if the platform is TILEmpower that needs
* special configuration for the PLX switch chip.
*/
int __write_once tile_plx_gen1;
static struct pci_controller controllers[TILE_NUM_PCIE];
static int num_controllers;
static struct pci_ops tile_cfg_ops;
/*
* We don't need to worry about the alignment of resources.
*/
resource_size_t pcibios_align_resource(void *data, const struct resource *res,
resource_size_t size, resource_size_t align)
{
return res->start;
}
EXPORT_SYMBOL(pcibios_align_resource);
/*
* Open a FD to the hypervisor PCI device.
*
* controller_id is the controller number, config type is 0 or 1 for
* config0 or config1 operations.
*/
static int __init tile_pcie_open(int controller_id, int config_type)
{
char filename[32];
int fd;
sprintf(filename, "pcie/%d/config%d", controller_id, config_type);
fd = hv_dev_open((HV_VirtAddr)filename, 0);
return fd;
}
/*
* Get the IRQ numbers from the HV and set up the handlers for them.
*/
static int __init tile_init_irqs(int controller_id,
struct pci_controller *controller)
{
char filename[32];
int fd;
int ret;
int x;
struct pcie_rc_config rc_config;
sprintf(filename, "pcie/%d/ctl", controller_id);
fd = hv_dev_open((HV_VirtAddr)filename, 0);
if (fd < 0) {
pr_err("PCI: hv_dev_open(%s) failed\n", filename);
return -1;
}
ret = hv_dev_pread(fd, 0, (HV_VirtAddr)(&rc_config),
sizeof(rc_config), PCIE_RC_CONFIG_MASK_OFF);
hv_dev_close(fd);
if (ret != sizeof(rc_config)) {
pr_err("PCI: wanted %zd bytes, got %d\n",
sizeof(rc_config), ret);
return -1;
}
/* Record irq_base so that we can map INTx to IRQ # later. */
controller->irq_base = rc_config.intr;
for (x = 0; x < 4; x++)
tile_irq_activate(rc_config.intr + x,
TILE_IRQ_HW_CLEAR);
if (rc_config.plx_gen1)
controller->plx_gen1 = 1;
return 0;
}
/*
* First initialization entry point, called from setup_arch().
*
* Find valid controllers and fill in pci_controller structs for each
* of them.
*
* Returns the number of controllers discovered.
*/
int __init tile_pci_init(void)
{
int i;
pr_info("PCI: Searching for controllers...\n");
/* Do any configuration we need before using the PCIe */
for (i = 0; i < TILE_NUM_PCIE; i++) {
int hv_cfg_fd0 = -1;
int hv_cfg_fd1 = -1;
int hv_mem_fd = -1;
char name[32];
struct pci_controller *controller;
/*
* Open the fd to the HV. If it fails then this
* device doesn't exist.
*/
hv_cfg_fd0 = tile_pcie_open(i, 0);
if (hv_cfg_fd0 < 0)
continue;
hv_cfg_fd1 = tile_pcie_open(i, 1);
if (hv_cfg_fd1 < 0) {
pr_err("PCI: Couldn't open config fd to HV "
"for controller %d\n", i);
goto err_cont;
}
sprintf(name, "pcie/%d/mem", i);
hv_mem_fd = hv_dev_open((HV_VirtAddr)name, 0);
if (hv_mem_fd < 0) {
pr_err("PCI: Could not open mem fd to HV!\n");
goto err_cont;
}
pr_info("PCI: Found PCI controller #%d\n", i);
controller = &controllers[num_controllers];
if (tile_init_irqs(i, controller)) {
pr_err("PCI: Could not initialize "
"IRQs, aborting.\n");
goto err_cont;
}
controller->index = num_controllers;
controller->hv_cfg_fd[0] = hv_cfg_fd0;
controller->hv_cfg_fd[1] = hv_cfg_fd1;
controller->hv_mem_fd = hv_mem_fd;
controller->first_busno = 0;
controller->last_busno = 0xff;
controller->ops = &tile_cfg_ops;
num_controllers++;
continue;
err_cont:
if (hv_cfg_fd0 >= 0)
hv_dev_close(hv_cfg_fd0);
if (hv_cfg_fd1 >= 0)
hv_dev_close(hv_cfg_fd1);
if (hv_mem_fd >= 0)
hv_dev_close(hv_mem_fd);
continue;
}
/*
* Before using the PCIe, see if we need to do any platform-specific
* configuration, such as the PLX switch Gen 1 issue on TILEmpower.
*/
for (i = 0; i < num_controllers; i++) {
struct pci_controller *controller = &controllers[i];
if (controller->plx_gen1)
tile_plx_gen1 = 1;
}
return num_controllers;
}
/*
* (pin - 1) converts from the PCI standard's [1:4] convention to
* a normal [0:3] range.
*/
static int tile_map_irq(struct pci_dev *dev, u8 slot, u8 pin)
{
struct pci_controller *controller =
(struct pci_controller *)dev->sysdata;
return (pin - 1) + controller->irq_base;
}
static void __init fixup_read_and_payload_sizes(void)
{
struct pci_dev *dev = NULL;
int smallest_max_payload = 0x1; /* Tile maxes out at 256 bytes. */
int max_read_size = 0x2; /* Limit to 512 byte reads. */
u16 new_values;
/* Scan for the smallest maximum payload size. */
while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
int pcie_caps_offset;
u32 devcap;
int max_payload;
pcie_caps_offset = pci_find_capability(dev, PCI_CAP_ID_EXP);
if (pcie_caps_offset == 0)
continue;
pci_read_config_dword(dev, pcie_caps_offset + PCI_EXP_DEVCAP,
&devcap);
max_payload = devcap & PCI_EXP_DEVCAP_PAYLOAD;
if (max_payload < smallest_max_payload)
smallest_max_payload = max_payload;
}
/* Now, set the max_payload_size for all devices to that value. */
new_values = (max_read_size << 12) | (smallest_max_payload << 5);
while ((dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev)) != NULL) {
int pcie_caps_offset;
u16 devctl;
pcie_caps_offset = pci_find_capability(dev, PCI_CAP_ID_EXP);
if (pcie_caps_offset == 0)
continue;
pci_read_config_word(dev, pcie_caps_offset + PCI_EXP_DEVCTL,
&devctl);
devctl &= ~(PCI_EXP_DEVCTL_PAYLOAD | PCI_EXP_DEVCTL_READRQ);
devctl |= new_values;
pci_write_config_word(dev, pcie_caps_offset + PCI_EXP_DEVCTL,
devctl);
}
}
/*
* Second PCI initialization entry point, called by subsys_initcall.
*
* The controllers have been set up by the time we get here, by a call to
* tile_pci_init.
*/
static int __init pcibios_init(void)
{
int i;
pr_info("PCI: Probing PCI hardware\n");
/*
* Delay a bit in case devices aren't ready. Some devices are
* known to require at least 20ms here, but we use a more
* conservative value.
*/
mdelay(250);
/* Scan all of the recorded PCI controllers. */
for (i = 0; i < num_controllers; i++) {
struct pci_controller *controller = &controllers[i];
struct pci_bus *bus;
pr_info("PCI: initializing controller #%d\n", i);
/*
* This comes from the generic Linux PCI driver.
*
* It reads the PCI tree for this bus into the Linux
* data structures.
*
* This is inlined in linux/pci.h and calls into
* pci_scan_bus_parented() in probe.c.
*/
bus = pci_scan_bus(0, controller->ops, controller);
controller->root_bus = bus;
controller->last_busno = bus->subordinate;
}
/* Do machine dependent PCI interrupt routing */
pci_fixup_irqs(pci_common_swizzle, tile_map_irq);
/*
* This comes from the generic Linux PCI driver.
*
* It allocates all of the resources (I/O memory, etc)
* associated with the devices read in above.
*/
pci_assign_unassigned_resources();
/* Configure the max_read_size and max_payload_size values. */
fixup_read_and_payload_sizes();
/* Record the I/O resources in the PCI controller structure. */
for (i = 0; i < num_controllers; i++) {
struct pci_bus *root_bus = controllers[i].root_bus;
struct pci_bus *next_bus;
struct pci_dev *dev;
list_for_each_entry(dev, &root_bus->devices, bus_list) {
/* Find the PCI host controller, ie. the 1st bridge. */
if ((dev->class >> 8) == PCI_CLASS_BRIDGE_PCI &&
(PCI_SLOT(dev->devfn) == 0)) {
next_bus = dev->subordinate;
controllers[i].mem_resources[0] =
*next_bus->resource[0];
controllers[i].mem_resources[1] =
*next_bus->resource[1];
controllers[i].mem_resources[2] =
*next_bus->resource[2];
break;
}
}
}
return 0;
}
subsys_initcall(pcibios_init);
/*
* No bus fixups needed.
*/
void __devinit pcibios_fixup_bus(struct pci_bus *bus)
{
/* Nothing needs to be done. */
}
/*
* This can be called from the generic PCI layer, but doesn't need to
* do anything.
*/
char __devinit *pcibios_setup(char *str)
{
/* Nothing needs to be done. */
return str;
}
/*
* This is called from the generic Linux layer.
*/
void __init pcibios_update_irq(struct pci_dev *dev, int irq)
{
pci_write_config_byte(dev, PCI_INTERRUPT_LINE, irq);
}
/*
* Enable memory and/or address decoding, as appropriate, for the
* device described by the 'dev' struct.
*
* This is called from the generic PCI layer, and can be called
* for bridges or endpoints.
*/
int pcibios_enable_device(struct pci_dev *dev, int mask)
{
u16 cmd, old_cmd;
u8 header_type;
int i;
struct resource *r;
pci_read_config_byte(dev, PCI_HEADER_TYPE, &header_type);
pci_read_config_word(dev, PCI_COMMAND, &cmd);
old_cmd = cmd;
if ((header_type & 0x7F) == PCI_HEADER_TYPE_BRIDGE) {
/*
* For bridges, we enable both memory and I/O decoding
* in call cases.
*/
cmd |= PCI_COMMAND_IO;
cmd |= PCI_COMMAND_MEMORY;
} else {
/*
* For endpoints, we enable memory and/or I/O decoding
* only if they have a memory resource of that type.
*/
for (i = 0; i < 6; i++) {
r = &dev->resource[i];
if (r->flags & IORESOURCE_UNSET) {
pr_err("PCI: Device %s not available "
"because of resource collisions\n",
pci_name(dev));
return -EINVAL;
}
if (r->flags & IORESOURCE_IO)
cmd |= PCI_COMMAND_IO;
if (r->flags & IORESOURCE_MEM)
cmd |= PCI_COMMAND_MEMORY;
}
}
/*
* We only write the command if it changed.
*/
if (cmd != old_cmd)
pci_write_config_word(dev, PCI_COMMAND, cmd);
return 0;
}
void __iomem *pci_iomap(struct pci_dev *dev, int bar, unsigned long max)
{
unsigned long start = pci_resource_start(dev, bar);
unsigned long len = pci_resource_len(dev, bar);
unsigned long flags = pci_resource_flags(dev, bar);
if (!len)
return NULL;
if (max && len > max)
len = max;
if (!(flags & IORESOURCE_MEM)) {
pr_info("PCI: Trying to map invalid resource %#lx\n", flags);
start = 0;
}
return (void __iomem *)start;
}
EXPORT_SYMBOL(pci_iomap);
/****************************************************************
*
* Tile PCI config space read/write routines
*
****************************************************************/
/*
* These are the normal read and write ops
* These are expanded with macros from pci_bus_read_config_byte() etc.
*
* devfn is the combined PCI slot & function.
*
* offset is in bytes, from the start of config space for the
* specified bus & slot.
*/
static int __devinit tile_cfg_read(struct pci_bus *bus,
unsigned int devfn,
int offset,
int size,
u32 *val)
{
struct pci_controller *controller = bus->sysdata;
int busnum = bus->number & 0xff;
int slot = (devfn >> 3) & 0x1f;
int function = devfn & 0x7;
u32 addr;
int config_mode = 1;
/*
* There is no bridge between the Tile and bus 0, so we
* use config0 to talk to bus 0.
*
* If we're talking to a bus other than zero then we
* must have found a bridge.
*/
if (busnum == 0) {
/*
* We fake an empty slot for (busnum == 0) && (slot > 0),
* since there is only one slot on bus 0.
*/
if (slot) {
*val = 0xFFFFFFFF;
return 0;
}
config_mode = 0;
}
addr = busnum << 20; /* Bus in 27:20 */
addr |= slot << 15; /* Slot (device) in 19:15 */
addr |= function << 12; /* Function is in 14:12 */
addr |= (offset & 0xFFF); /* byte address in 0:11 */
return hv_dev_pread(controller->hv_cfg_fd[config_mode], 0,
(HV_VirtAddr)(val), size, addr);
}
/*
* See tile_cfg_read() for relevent comments.
* Note that "val" is the value to write, not a pointer to that value.
*/
static int __devinit tile_cfg_write(struct pci_bus *bus,
unsigned int devfn,
int offset,
int size,
u32 val)
{
struct pci_controller *controller = bus->sysdata;
int busnum = bus->number & 0xff;
int slot = (devfn >> 3) & 0x1f;
int function = devfn & 0x7;
u32 addr;
int config_mode = 1;
HV_VirtAddr valp = (HV_VirtAddr)&val;
/*
* For bus 0 slot 0 we use config 0 accesses.
*/
if (busnum == 0) {
/*
* We fake an empty slot for (busnum == 0) && (slot > 0),
* since there is only one slot on bus 0.
*/
if (slot)
return 0;
config_mode = 0;
}
addr = busnum << 20; /* Bus in 27:20 */
addr |= slot << 15; /* Slot (device) in 19:15 */
addr |= function << 12; /* Function is in 14:12 */
addr |= (offset & 0xFFF); /* byte address in 0:11 */
#ifdef __BIG_ENDIAN
/* Point to the correct part of the 32-bit "val". */
valp += 4 - size;
#endif
return hv_dev_pwrite(controller->hv_cfg_fd[config_mode], 0,
valp, size, addr);
}
static struct pci_ops tile_cfg_ops = {
.read = tile_cfg_read,
.write = tile_cfg_write,
};
/*
* In the following, each PCI controller's mem_resources[1]
* represents its (non-prefetchable) PCI memory resource.
* mem_resources[0] and mem_resources[2] refer to its PCI I/O and
* prefetchable PCI memory resources, respectively.
* For more details, see pci_setup_bridge() in setup-bus.c.
* By comparing the target PCI memory address against the
* end address of controller 0, we can determine the controller
* that should accept the PCI memory access.
*/
#define TILE_READ(size, type) \
type _tile_read##size(unsigned long addr) \
{ \
type val; \
int idx = 0; \
if (addr > controllers[0].mem_resources[1].end && \
addr > controllers[0].mem_resources[2].end) \
idx = 1; \
if (hv_dev_pread(controllers[idx].hv_mem_fd, 0, \
(HV_VirtAddr)(&val), sizeof(type), addr)) \
pr_err("PCI: read %zd bytes at 0x%lX failed\n", \
sizeof(type), addr); \
return val; \
} \
EXPORT_SYMBOL(_tile_read##size)
TILE_READ(b, u8);
TILE_READ(w, u16);
TILE_READ(l, u32);
TILE_READ(q, u64);
#define TILE_WRITE(size, type) \
void _tile_write##size(type val, unsigned long addr) \
{ \
int idx = 0; \
if (addr > controllers[0].mem_resources[1].end && \
addr > controllers[0].mem_resources[2].end) \
idx = 1; \
if (hv_dev_pwrite(controllers[idx].hv_mem_fd, 0, \
(HV_VirtAddr)(&val), sizeof(type), addr)) \
pr_err("PCI: write %zd bytes at 0x%lX failed\n", \
sizeof(type), addr); \
} \
EXPORT_SYMBOL(_tile_write##size)
TILE_WRITE(b, u8);
TILE_WRITE(w, u16);
TILE_WRITE(l, u32);
TILE_WRITE(q, u64);

2
arch/tile/kernel/setup.c
View file

@ -840,7 +840,7 @@ static int __init topology_init(void)
for_each_online_node(i)
register_one_node(i);
for_each_present_cpu(i)
for (i = 0; i < smp_height * smp_width; ++i)
register_cpu(&cpu_devices[i], i);
return 0;

35
arch/tile/lib/memchr_32.c
View file

@ -18,12 +18,24 @@
void *memchr(const void *s, int c, size_t n)
{
const uint32_t *last_word_ptr;
const uint32_t *p;
const char *last_byte_ptr;
uintptr_t s_int;
uint32_t goal, before_mask, v, bits;
char *ret;
if (__builtin_expect(n == 0, 0)) {
/* Don't dereference any memory if the array is empty. */
return NULL;
}
/* Get an aligned pointer. */
const uintptr_t s_int = (uintptr_t) s;
const uint32_t *p = (const uint32_t *)(s_int & -4);
s_int = (uintptr_t) s;
p = (const uint32_t *)(s_int & -4);
/* Create four copies of the byte for which we are looking. */
const uint32_t goal = 0x01010101 * (uint8_t) c;
goal = 0x01010101 * (uint8_t) c;
/* Read the first word, but munge it so that bytes before the array
* will not match goal.
@ -31,23 +43,14 @@ void *memchr(const void *s, int c, size_t n)
* Note that this shift count expression works because we know
* shift counts are taken mod 32.
*/
const uint32_t before_mask = (1 << (s_int << 3)) - 1;
uint32_t v = (*p | before_mask) ^ (goal & before_mask);
before_mask = (1 << (s_int << 3)) - 1;
v = (*p | before_mask) ^ (goal & before_mask);
/* Compute the address of the last byte. */
const char *const last_byte_ptr = (const char *)s + n - 1;
last_byte_ptr = (const char *)s + n - 1;
/* Compute the address of the word containing the last byte. */
const uint32_t *const last_word_ptr =
(const uint32_t *)((uintptr_t) last_byte_ptr & -4);
uint32_t bits;
char *ret;
if (__builtin_expect(n == 0, 0)) {
/* Don't dereference any memory if the array is empty. */
return NULL;
}
last_word_ptr = (const uint32_t *)((uintptr_t) last_byte_ptr & -4);
while ((bits = __insn_seqb(v, goal)) == 0) {
if (__builtin_expect(p == last_word_ptr, 0)) {

31
arch/tile/lib/spinlock_32.c
View file

@ -167,23 +167,30 @@ void arch_write_lock_slow(arch_rwlock_t *rwlock, u32 val)
* when we compare them.
*/
u32 my_ticket_;
/* Take out the next ticket; this will also stop would-be readers. */
if (val & 1)
val = get_rwlock(rwlock);
rwlock->lock = __insn_addb(val, 1 << WR_NEXT_SHIFT);
/* Extract my ticket value from the original word. */
my_ticket_ = val >> WR_NEXT_SHIFT;
u32 iterations = 0;
/*
* Wait until the "current" field matches our ticket, and
* there are no remaining readers.
* Wait until there are no readers, then bump up the next
* field and capture the ticket value.
*/
for (;;) {
if (!(val & 1)) {
if ((val >> RD_COUNT_SHIFT) == 0)
break;
rwlock->lock = val;
}
delay_backoff(iterations++);
val = __insn_tns((int *)&rwlock->lock);
}
/* Take out the next ticket and extract my ticket value. */
rwlock->lock = __insn_addb(val, 1 << WR_NEXT_SHIFT);
my_ticket_ = val >> WR_NEXT_SHIFT;
/* Wait until the "current" field matches our ticket. */
for (;;) {
u32 curr_ = val >> WR_CURR_SHIFT;
u32 readers = val >> RD_COUNT_SHIFT;
u32 delta = ((my_ticket_ - curr_) & WR_MASK) + !!readers;
u32 delta = ((my_ticket_ - curr_) & WR_MASK);
if (likely(delta == 0))
break;

5
arch/um/drivers/line.c
View file

@ -727,6 +727,9 @@ struct winch {
static void free_winch(struct winch *winch, int free_irq_ok)
{
if (free_irq_ok)
free_irq(WINCH_IRQ, winch);
list_del(&winch->list);
if (winch->pid != -1)
@ -735,8 +738,6 @@ static void free_winch(struct winch *winch, int free_irq_ok)
os_close_file(winch->fd);
if (winch->stack != 0)
free_stack(winch->stack, 0);
if (free_irq_ok)
free_irq(WINCH_IRQ, winch);
kfree(winch);
}

2
arch/x86/Kconfig
View file

@ -21,7 +21,7 @@ config X86
select HAVE_UNSTABLE_SCHED_CLOCK
select HAVE_IDE
select HAVE_OPROFILE
select HAVE_PERF_EVENTS if (!M386 && !M486)
select HAVE_PERF_EVENTS
select HAVE_IRQ_WORK
select HAVE_IOREMAP_PROT
select HAVE_KPROBES

4
arch/x86/include/asm/fixmap.h
View file

@ -216,8 +216,8 @@ static inline unsigned long virt_to_fix(const unsigned long vaddr)
}
/* Return an pointer with offset calculated */
static inline unsigned long __set_fixmap_offset(enum fixed_addresses idx,
phys_addr_t phys, pgprot_t flags)
static __always_inline unsigned long
__set_fixmap_offset(enum fixed_addresses idx, phys_addr_t phys, pgprot_t flags)
{
__set_fixmap(idx, phys, flags);
return fix_to_virt(idx) + (phys & (PAGE_SIZE - 1));

2
arch/x86/include/asm/msr-index.h
View file

@ -128,7 +128,7 @@
#define FAM10H_MMIO_CONF_ENABLE (1<<0)
#define FAM10H_MMIO_CONF_BUSRANGE_MASK 0xf
#define FAM10H_MMIO_CONF_BUSRANGE_SHIFT 2
#define FAM10H_MMIO_CONF_BASE_MASK 0xfffffff
#define FAM10H_MMIO_CONF_BASE_MASK 0xfffffffULL
#define FAM10H_MMIO_CONF_BASE_SHIFT 20
#define MSR_FAM10H_NODE_ID 0xc001100c

10
arch/x86/include/asm/paravirt.h
View file

@ -824,27 +824,27 @@ static __always_inline void arch_spin_unlock(struct arch_spinlock *lock)
#define __PV_IS_CALLEE_SAVE(func) \
((struct paravirt_callee_save) { func })
static inline unsigned long arch_local_save_flags(void)
static inline notrace unsigned long arch_local_save_flags(void)
{
return PVOP_CALLEE0(unsigned long, pv_irq_ops.save_fl);
}
static inline void arch_local_irq_restore(unsigned long f)
static inline notrace void arch_local_irq_restore(unsigned long f)
{
PVOP_VCALLEE1(pv_irq_ops.restore_fl, f);
}
static inline void arch_local_irq_disable(void)
static inline notrace void arch_local_irq_disable(void)
{
PVOP_VCALLEE0(pv_irq_ops.irq_disable);
}
static inline void arch_local_irq_enable(void)
static inline notrace void arch_local_irq_enable(void)
{
PVOP_VCALLEE0(pv_irq_ops.irq_enable);
}
static inline unsigned long arch_local_irq_save(void)
static inline notrace unsigned long arch_local_irq_save(void)
{
unsigned long f;

4
arch/x86/include/asm/uv/uv_hub.h
View file

@ -199,6 +199,8 @@ union uvh_apicid {
#define UVH_APICID 0x002D0E00L
#define UV_APIC_PNODE_SHIFT 6
#define UV_APICID_HIBIT_MASK 0xffff0000
/* Local Bus from cpu's perspective */
#define LOCAL_BUS_BASE 0x1c00000
#define LOCAL_BUS_SIZE (4 * 1024 * 1024)
@ -491,8 +493,10 @@ static inline void uv_set_cpu_scir_bits(int cpu, unsigned char value)
}
}
extern unsigned int uv_apicid_hibits;
static unsigned long uv_hub_ipi_value(int apicid, int vector, int mode)
{
apicid |= uv_apicid_hibits;
return (1UL << UVH_IPI_INT_SEND_SHFT) |
((apicid) << UVH_IPI_INT_APIC_ID_SHFT) |
(mode << UVH_IPI_INT_DELIVERY_MODE_SHFT) |

19
arch/x86/include/asm/uv/uv_mmrs.h
View file

@ -5,7 +5,7 @@
*
* SGI UV MMR definitions
*
* Copyright (C) 2007-2008 Silicon Graphics, Inc. All rights reserved.
* Copyright (C) 2007-2010 Silicon Graphics, Inc. All rights reserved.
*/
#ifndef _ASM_X86_UV_UV_MMRS_H
@ -753,6 +753,23 @@ union uvh_lb_bau_sb_descriptor_base_u {
} s;
};
/* ========================================================================= */
/* UVH_LB_TARGET_PHYSICAL_APIC_ID_MASK */
/* ========================================================================= */
#define UVH_LB_TARGET_PHYSICAL_APIC_ID_MASK 0x320130UL
#define UVH_LB_TARGET_PHYSICAL_APIC_ID_MASK_32 0x009f0
#define UVH_LB_TARGET_PHYSICAL_APIC_ID_MASK_BIT_ENABLES_SHFT 0
#define UVH_LB_TARGET_PHYSICAL_APIC_ID_MASK_BIT_ENABLES_MASK 0x00000000ffffffffUL
union uvh_lb_target_physical_apic_id_mask_u {
unsigned long v;
struct uvh_lb_target_physical_apic_id_mask_s {
unsigned long bit_enables : 32; /* RW */
unsigned long rsvd_32_63 : 32; /* */
} s;
};
/* ========================================================================= */
/* UVH_NODE_ID */
/* ========================================================================= */

6
arch/x86/include/asm/xen/interface.h
View file

@ -61,9 +61,9 @@ DEFINE_GUEST_HANDLE(void);
#define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START)
#endif
#ifndef machine_to_phys_mapping
#define machine_to_phys_mapping ((unsigned long *)HYPERVISOR_VIRT_START)
#endif
#define MACH2PHYS_VIRT_START mk_unsigned_long(__MACH2PHYS_VIRT_START)
#define MACH2PHYS_VIRT_END mk_unsigned_long(__MACH2PHYS_VIRT_END)
#define MACH2PHYS_NR_ENTRIES ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>__MACH2PHYS_SHIFT)
/* Maximum number of virtual CPUs in multi-processor guests. */
#define MAX_VIRT_CPUS 32

5
arch/x86/include/asm/xen/interface_32.h
View file

@ -32,6 +32,11 @@
/* And the trap vector is... */
#define TRAP_INSTR "int $0x82"
#define __MACH2PHYS_VIRT_START 0xF5800000
#define __MACH2PHYS_VIRT_END 0xF6800000
#define __MACH2PHYS_SHIFT 2
/*
* Virtual addresses beyond this are not modifiable by guest OSes. The
* machine->physical mapping table starts at this address, read-only.

13
arch/x86/include/asm/xen/interface_64.h
View file

@ -39,18 +39,7 @@
#define __HYPERVISOR_VIRT_END 0xFFFF880000000000
#define __MACH2PHYS_VIRT_START 0xFFFF800000000000
#define __MACH2PHYS_VIRT_END 0xFFFF804000000000
#ifndef HYPERVISOR_VIRT_START
#define HYPERVISOR_VIRT_START mk_unsigned_long(__HYPERVISOR_VIRT_START)
#define HYPERVISOR_VIRT_END mk_unsigned_long(__HYPERVISOR_VIRT_END)
#endif
#define MACH2PHYS_VIRT_START mk_unsigned_long(__MACH2PHYS_VIRT_START)
#define MACH2PHYS_VIRT_END mk_unsigned_long(__MACH2PHYS_VIRT_END)
#define MACH2PHYS_NR_ENTRIES ((MACH2PHYS_VIRT_END-MACH2PHYS_VIRT_START)>>3)
#ifndef machine_to_phys_mapping
#define machine_to_phys_mapping ((unsigned long *)HYPERVISOR_VIRT_START)
#endif
#define __MACH2PHYS_SHIFT 3
/*
* int HYPERVISOR_set_segment_base(unsigned int which, unsigned long base)

7
arch/x86/include/asm/xen/page.h
View file

@ -5,6 +5,7 @@
#include <linux/types.h>
#include <linux/spinlock.h>
#include <linux/pfn.h>
#include <linux/mm.h>
#include <asm/uaccess.h>
#include <asm/page.h>
@ -35,6 +36,8 @@ typedef struct xpaddr {
#define MAX_DOMAIN_PAGES \
((unsigned long)((u64)CONFIG_XEN_MAX_DOMAIN_MEMORY * 1024 * 1024 * 1024 / PAGE_SIZE))
extern unsigned long *machine_to_phys_mapping;
extern unsigned int machine_to_phys_order;
extern unsigned long get_phys_to_machine(unsigned long pfn);
extern bool set_phys_to_machine(unsigned long pfn, unsigned long mfn);
@ -69,10 +72,8 @@ static inline unsigned long mfn_to_pfn(unsigned long mfn)
if (xen_feature(XENFEAT_auto_translated_physmap))
return mfn;
#if 0
if (unlikely((mfn >> machine_to_phys_order) != 0))
return max_mapnr;
#endif
return ~0;
pfn = 0;
/*

7
arch/x86/kernel/apic/hw_nmi.c
View file

@ -17,15 +17,16 @@
#include <linux/nmi.h>
#include <linux/module.h>
/* For reliability, we're prepared to waste bits here. */
static DECLARE_BITMAP(backtrace_mask, NR_CPUS) __read_mostly;
u64 hw_nmi_get_sample_period(void)
{
return (u64)(cpu_khz) * 1000 * 60;
}
#ifdef ARCH_HAS_NMI_WATCHDOG
/* For reliability, we're prepared to waste bits here. */
static DECLARE_BITMAP(backtrace_mask, NR_CPUS) __read_mostly;
void arch_trigger_all_cpu_backtrace(void)
{
int i;

25
arch/x86/kernel/apic/x2apic_uv_x.c
View file

@ -44,6 +44,8 @@ static u64 gru_start_paddr, gru_end_paddr;
static union uvh_apicid uvh_apicid;
int uv_min_hub_revision_id;
EXPORT_SYMBOL_GPL(uv_min_hub_revision_id);
unsigned int uv_apicid_hibits;
EXPORT_SYMBOL_GPL(uv_apicid_hibits);
static DEFINE_SPINLOCK(uv_nmi_lock);
static inline bool is_GRU_range(u64 start, u64 end)
@ -85,6 +87,23 @@ static void __init early_get_apic_pnode_shift(void)
uvh_apicid.s.pnode_shift = UV_APIC_PNODE_SHIFT;
}
/*
* Add an extra bit as dictated by bios to the destination apicid of
* interrupts potentially passing through the UV HUB. This prevents
* a deadlock between interrupts and IO port operations.
*/
static void __init uv_set_apicid_hibit(void)
{
union uvh_lb_target_physical_apic_id_mask_u apicid_mask;
unsigned long *mmr;
mmr = early_ioremap(UV_LOCAL_MMR_BASE |
UVH_LB_TARGET_PHYSICAL_APIC_ID_MASK, sizeof(*mmr));
apicid_mask.v = *mmr;
early_iounmap(mmr, sizeof(*mmr));
uv_apicid_hibits = apicid_mask.s.bit_enables & UV_APICID_HIBIT_MASK;
}
static int __init uv_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
{
int nodeid;
@ -102,6 +121,7 @@ static int __init uv_acpi_madt_oem_check(char *oem_id, char *oem_table_id)
__get_cpu_var(x2apic_extra_bits) =
nodeid << (uvh_apicid.s.pnode_shift - 1);
uv_system_type = UV_NON_UNIQUE_APIC;
uv_set_apicid_hibit();
return 1;
}
}
@ -155,6 +175,7 @@ static int __cpuinit uv_wakeup_secondary(int phys_apicid, unsigned long start_ri
int pnode;
pnode = uv_apicid_to_pnode(phys_apicid);
phys_apicid |= uv_apicid_hibits;
val = (1UL << UVH_IPI_INT_SEND_SHFT) |
(phys_apicid << UVH_IPI_INT_APIC_ID_SHFT) |
((start_rip << UVH_IPI_INT_VECTOR_SHFT) >> 12) |
@ -236,7 +257,7 @@ static unsigned int uv_cpu_mask_to_apicid(const struct cpumask *cpumask)
int cpu = cpumask_first(cpumask);
if ((unsigned)cpu < nr_cpu_ids)
return per_cpu(x86_cpu_to_apicid, cpu);
return per_cpu(x86_cpu_to_apicid, cpu) | uv_apicid_hibits;
else
return BAD_APICID;
}
@ -255,7 +276,7 @@ uv_cpu_mask_to_apicid_and(const struct cpumask *cpumask,
if (cpumask_test_cpu(cpu, cpu_online_mask))
break;
}
return per_cpu(x86_cpu_to_apicid, cpu);
return per_cpu(x86_cpu_to_apicid, cpu) | uv_apicid_hibits;
}
static unsigned int x2apic_get_apic_id(unsigned long x)

20
arch/x86/kernel/cpu/perf_event.c
View file

@ -381,6 +381,20 @@ static void release_pmc_hardware(void) {}
#endif
static bool check_hw_exists(void)
{
u64 val, val_new = 0;
int ret = 0;
val = 0xabcdUL;
ret |= checking_wrmsrl(x86_pmu.perfctr, val);
ret |= rdmsrl_safe(x86_pmu.perfctr, &val_new);
if (ret || val != val_new)
return false;
return true;
}
static void reserve_ds_buffers(void);
static void release_ds_buffers(void);
@ -1372,6 +1386,12 @@ void __init init_hw_perf_events(void)
pmu_check_apic();
/* sanity check that the hardware exists or is emulated */
if (!check_hw_exists()) {
pr_cont("Broken PMU hardware detected, software events only.\n");
return;
}
pr_cont("%s PMU driver.\n", x86_pmu.name);
if (x86_pmu.quirks)

2
arch/x86/kernel/entry_32.S
View file

@ -395,7 +395,7 @@ sysenter_past_esp:
* A tiny bit of offset fixup is necessary - 4*4 means the 4 words
* pushed above; +8 corresponds to copy_thread's esp0 setting.
*/
pushl_cfi (TI_sysenter_return-THREAD_SIZE_asm+8+4*4)(%esp)
pushl_cfi ((TI_sysenter_return)-THREAD_SIZE_asm+8+4*4)(%esp)
CFI_REL_OFFSET eip, 0
pushl_cfi %eax

2
arch/x86/kernel/entry_64.S
View file

@ -295,6 +295,7 @@ ENDPROC(native_usergs_sysret64)
.endm
/* save partial stack frame */
.pushsection .kprobes.text, "ax"
ENTRY(save_args)
XCPT_FRAME
cld
@ -334,6 +335,7 @@ ENTRY(save_args)
ret
CFI_ENDPROC
END(save_args)
.popsection
ENTRY(save_rest)
PARTIAL_FRAME 1 REST_SKIP+8

4
arch/x86/kernel/hw_breakpoint.c
View file

@ -433,6 +433,10 @@ static int __kprobes hw_breakpoint_handler(struct die_args *args)
dr6_p = (unsigned long *)ERR_PTR(args->err);
dr6 = *dr6_p;
/* If it's a single step, TRAP bits are random */
if (dr6 & DR_STEP)
return NOTIFY_DONE;
/* Do an early return if no trap bits are set in DR6 */
if ((dr6 & DR_TRAP_BITS) == 0)
return NOTIFY_DONE;

64
arch/x86/kernel/mmconf-fam10h_64.c
View file

@ -25,7 +25,6 @@ struct pci_hostbridge_probe {
};
static u64 __cpuinitdata fam10h_pci_mmconf_base;
static int __cpuinitdata fam10h_pci_mmconf_base_status;
static struct pci_hostbridge_probe pci_probes[] __cpuinitdata = {
{ 0, 0x18, PCI_VENDOR_ID_AMD, 0x1200 },
@ -44,10 +43,12 @@ static int __cpuinit cmp_range(const void *x1, const void *x2)
return start1 - start2;
}
/*[47:0] */
/* need to avoid (0xfd<<32) and (0xfe<<32), ht used space */
#define MMCONF_UNIT (1ULL << FAM10H_MMIO_CONF_BASE_SHIFT)
#define MMCONF_MASK (~(MMCONF_UNIT - 1))
#define MMCONF_SIZE (MMCONF_UNIT << 8)
/* need to avoid (0xfd<<32), (0xfe<<32), and (0xff<<32), ht used space */
#define FAM10H_PCI_MMCONF_BASE (0xfcULL<<32)
#define BASE_VALID(b) ((b != (0xfdULL << 32)) && (b != (0xfeULL << 32)))
#define BASE_VALID(b) ((b) + MMCONF_SIZE <= (0xfdULL<<32) || (b) >= (1ULL<<40))
static void __cpuinit get_fam10h_pci_mmconf_base(void)
{
int i;
@ -64,12 +65,11 @@ static void __cpuinit get_fam10h_pci_mmconf_base(void)
struct range range[8];
/* only try to get setting from BSP */
/* -1 or 1 */
if (fam10h_pci_mmconf_base_status)
if (fam10h_pci_mmconf_base)
return;
if (!early_pci_allowed())
goto fail;
return;
found = 0;
for (i = 0; i < ARRAY_SIZE(pci_probes); i++) {
@ -91,7 +91,7 @@ static void __cpuinit get_fam10h_pci_mmconf_base(void)
}
if (!found)
goto fail;
return;
/* SYS_CFG */
address = MSR_K8_SYSCFG;
@ -99,16 +99,16 @@ static void __cpuinit get_fam10h_pci_mmconf_base(void)
/* TOP_MEM2 is not enabled? */
if (!(val & (1<<21))) {
tom2 = 0;
tom2 = 1ULL << 32;
} else {
/* TOP_MEM2 */
address = MSR_K8_TOP_MEM2;
rdmsrl(address, val);
tom2 = val & (0xffffULL<<32);
tom2 = max(val & 0xffffff800000ULL, 1ULL << 32);
}
if (base <= tom2)
base = tom2 + (1ULL<<32);
base = (tom2 + 2 * MMCONF_UNIT - 1) & MMCONF_MASK;
/*
* need to check if the range is in the high mmio range that is
@ -123,11 +123,11 @@ static void __cpuinit get_fam10h_pci_mmconf_base(void)
if (!(reg & 3))
continue;
start = (((u64)reg) << 8) & (0xffULL << 32); /* 39:16 on 31:8*/
start = (u64)(reg & 0xffffff00) << 8; /* 39:16 on 31:8*/
reg = read_pci_config(bus, slot, 1, 0x84 + (i << 3));
end = (((u64)reg) << 8) & (0xffULL << 32); /* 39:16 on 31:8*/
end = ((u64)(reg & 0xffffff00) << 8) | 0xffff; /* 39:16 on 31:8*/
if (!end)
if (end < tom2)
continue;
range[hi_mmio_num].start = start;
@ -143,32 +143,27 @@ static void __cpuinit get_fam10h_pci_mmconf_base(void)
if (range[hi_mmio_num - 1].end < base)
goto out;
if (range[0].start > base)
if (range[0].start > base + MMCONF_SIZE)
goto out;
/* need to find one window */
base = range[0].start - (1ULL << 32);
base = (range[0].start & MMCONF_MASK) - MMCONF_UNIT;
if ((base > tom2) && BASE_VALID(base))
goto out;
base = range[hi_mmio_num - 1].end + (1ULL << 32);
if ((base > tom2) && BASE_VALID(base))
base = (range[hi_mmio_num - 1].end + MMCONF_UNIT) & MMCONF_MASK;
if (BASE_VALID(base))
goto out;
/* need to find window between ranges */
if (hi_mmio_num > 1)
for (i = 0; i < hi_mmio_num - 1; i++) {
if (range[i + 1].start > (range[i].end + (1ULL << 32))) {
base = range[i].end + (1ULL << 32);
if ((base > tom2) && BASE_VALID(base))
goto out;
}
for (i = 1; i < hi_mmio_num; i++) {
base = (range[i - 1].end + MMCONF_UNIT) & MMCONF_MASK;
val = range[i].start & MMCONF_MASK;
if (val >= base + MMCONF_SIZE && BASE_VALID(base))
goto out;
}
fail:
fam10h_pci_mmconf_base_status = -1;
return;
out:
fam10h_pci_mmconf_base = base;
fam10h_pci_mmconf_base_status = 1;
}
void __cpuinit fam10h_check_enable_mmcfg(void)
@ -190,11 +185,10 @@ void __cpuinit fam10h_check_enable_mmcfg(void)
/* only trust the one handle 256 buses, if acpi=off */
if (!acpi_pci_disabled || busnbits >= 8) {
u64 base;
base = val & (0xffffULL << 32);
if (fam10h_pci_mmconf_base_status <= 0) {
u64 base = val & MMCONF_MASK;
if (!fam10h_pci_mmconf_base) {
fam10h_pci_mmconf_base = base;
fam10h_pci_mmconf_base_status = 1;
return;
} else if (fam10h_pci_mmconf_base == base)
return;
@ -206,8 +200,10 @@ void __cpuinit fam10h_check_enable_mmcfg(void)
* with 256 buses
*/
get_fam10h_pci_mmconf_base();
if (fam10h_pci_mmconf_base_status <= 0)
if (!fam10h_pci_mmconf_base) {
pci_probe &= ~PCI_CHECK_ENABLE_AMD_MMCONF;
return;
}
printk(KERN_INFO "Enable MMCONFIG on AMD Family 10h\n");
val &= ~((FAM10H_MMIO_CONF_BASE_MASK<<FAM10H_MMIO_CONF_BASE_SHIFT) |

5
arch/x86/mm/tlb.c
View file

@ -223,7 +223,7 @@ void native_flush_tlb_others(const struct cpumask *cpumask,
static void __cpuinit calculate_tlb_offset(void)
{
int cpu, node, nr_node_vecs;
int cpu, node, nr_node_vecs, idx = 0;
/*
* we are changing tlb_vector_offset for each CPU in runtime, but this
* will not cause inconsistency, as the write is atomic under X86. we
@ -239,7 +239,7 @@ static void __cpuinit calculate_tlb_offset(void)
nr_node_vecs = NUM_INVALIDATE_TLB_VECTORS/nr_online_nodes;
for_each_online_node(node) {
int node_offset = (node % NUM_INVALIDATE_TLB_VECTORS) *
int node_offset = (idx % NUM_INVALIDATE_TLB_VECTORS) *
nr_node_vecs;
int cpu_offset = 0;
for_each_cpu(cpu, cpumask_of_node(node)) {
@ -248,6 +248,7 @@ static void __cpuinit calculate_tlb_offset(void)
cpu_offset++;
cpu_offset = cpu_offset % nr_node_vecs;
}
idx++;
}
}

2
arch/x86/platform/uv/tlb_uv.c
View file

@ -1455,7 +1455,7 @@ static void __init uv_init_uvhub(int uvhub, int vector)
* the below initialization can't be in firmware because the
* messaging IRQ will be determined by the OS
*/
apicid = uvhub_to_first_apicid(uvhub);
apicid = uvhub_to_first_apicid(uvhub) | uv_apicid_hibits;
uv_write_global_mmr64(pnode, UVH_BAU_DATA_CONFIG,
((apicid << 32) | vector));
}

4
arch/x86/platform/uv/uv_time.c
View file

@ -89,6 +89,7 @@ static void uv_rtc_send_IPI(int cpu)
apicid = cpu_physical_id(cpu);
pnode = uv_apicid_to_pnode(apicid);
apicid |= uv_apicid_hibits;
val = (1UL << UVH_IPI_INT_SEND_SHFT) |
(apicid << UVH_IPI_INT_APIC_ID_SHFT) |
(X86_PLATFORM_IPI_VECTOR << UVH_IPI_INT_VECTOR_SHFT);
@ -107,6 +108,7 @@ static int uv_intr_pending(int pnode)
static int uv_setup_intr(int cpu, u64 expires)
{
u64 val;
unsigned long apicid = cpu_physical_id(cpu) | uv_apicid_hibits;
int pnode = uv_cpu_to_pnode(cpu);
uv_write_global_mmr64(pnode, UVH_RTC1_INT_CONFIG,
@ -117,7 +119,7 @@ static int uv_setup_intr(int cpu, u64 expires)
UVH_EVENT_OCCURRED0_RTC1_MASK);
val = (X86_PLATFORM_IPI_VECTOR << UVH_RTC1_INT_CONFIG_VECTOR_SHFT) |
((u64)cpu_physical_id(cpu) << UVH_RTC1_INT_CONFIG_APIC_ID_SHFT);
((u64)apicid << UVH_RTC1_INT_CONFIG_APIC_ID_SHFT);
/* Set configuration */
uv_write_global_mmr64(pnode, UVH_RTC1_INT_CONFIG, val);

21
arch/x86/xen/enlighten.c
View file

@ -75,6 +75,11 @@ DEFINE_PER_CPU(struct vcpu_info, xen_vcpu_info);
enum xen_domain_type xen_domain_type = XEN_NATIVE;
EXPORT_SYMBOL_GPL(xen_domain_type);
unsigned long *machine_to_phys_mapping = (void *)MACH2PHYS_VIRT_START;
EXPORT_SYMBOL(machine_to_phys_mapping);
unsigned int machine_to_phys_order;
EXPORT_SYMBOL(machine_to_phys_order);
struct start_info *xen_start_info;
EXPORT_SYMBOL_GPL(xen_start_info);
@ -1090,6 +1095,8 @@ static void __init xen_setup_stackprotector(void)
/* First C function to be called on Xen boot */
asmlinkage void __init xen_start_kernel(void)
{
struct physdev_set_iopl set_iopl;
int rc;
pgd_t *pgd;
if (!xen_start_info)
@ -1097,6 +1104,8 @@ asmlinkage void __init xen_start_kernel(void)
xen_domain_type = XEN_PV_DOMAIN;
xen_setup_machphys_mapping();
/* Install Xen paravirt ops */
pv_info = xen_info;
pv_init_ops = xen_init_ops;
@ -1191,8 +1200,6 @@ asmlinkage void __init xen_start_kernel(void)
/* Allocate and initialize top and mid mfn levels for p2m structure */
xen_build_mfn_list_list();
init_mm.pgd = pgd;
/* keep using Xen gdt for now; no urgent need to change it */
#ifdef CONFIG_X86_32
@ -1202,10 +1209,18 @@ asmlinkage void __init xen_start_kernel(void)
#else
pv_info.kernel_rpl = 0;
#endif
/* set the limit of our address space */
xen_reserve_top();
/* We used to do this in xen_arch_setup, but that is too late on AMD
* were early_cpu_init (run before ->arch_setup()) calls early_amd_init
* which pokes 0xcf8 port.
*/
set_iopl.iopl = 1;
rc = HYPERVISOR_physdev_op(PHYSDEVOP_set_iopl, &set_iopl);
if (rc != 0)
xen_raw_printk("physdev_op failed %d\n", rc);
#ifdef CONFIG_X86_32
/* set up basic CPUID stuff */
cpu_detect(&new_cpu_data);

Some files were not shown because too many files have changed in this diff Show more