aboutsummaryrefslogtreecommitdiffstats
path: root/drivers/platform/x86/msi-laptop.c
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/platform/x86/msi-laptop.c')
-rw-r--r--drivers/platform/x86/msi-laptop.c437
1 files changed, 437 insertions, 0 deletions
diff --git a/drivers/platform/x86/msi-laptop.c b/drivers/platform/x86/msi-laptop.c
new file mode 100644
index 000000000000..759763d18e4c
--- /dev/null
+++ b/drivers/platform/x86/msi-laptop.c
@@ -0,0 +1,437 @@
1/*-*-linux-c-*-*/
2
3/*
4 Copyright (C) 2006 Lennart Poettering <mzxreary (at) 0pointer (dot) de>
5
6 This program is free software; you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 2 of the License, or
9 (at your option) any later version.
10
11 This program is distributed in the hope that it will be useful, but
12 WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 General Public License for more details.
15
16 You should have received a copy of the GNU General Public License
17 along with this program; if not, write to the Free Software
18 Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
19 02110-1301, USA.
20 */
21
22/*
23 * msi-laptop.c - MSI S270 laptop support. This laptop is sold under
24 * various brands, including "Cytron/TCM/Medion/Tchibo MD96100".
25 *
26 * Driver also supports S271, S420 models.
27 *
28 * This driver exports a few files in /sys/devices/platform/msi-laptop-pf/:
29 *
30 * lcd_level - Screen brightness: contains a single integer in the
31 * range 0..8. (rw)
32 *
33 * auto_brightness - Enable automatic brightness control: contains
34 * either 0 or 1. If set to 1 the hardware adjusts the screen
35 * brightness automatically when the power cord is
36 * plugged/unplugged. (rw)
37 *
38 * wlan - WLAN subsystem enabled: contains either 0 or 1. (ro)
39 *
40 * bluetooth - Bluetooth subsystem enabled: contains either 0 or 1
41 * Please note that this file is constantly 0 if no Bluetooth
42 * hardware is available. (ro)
43 *
44 * In addition to these platform device attributes the driver
45 * registers itself in the Linux backlight control subsystem and is
46 * available to userspace under /sys/class/backlight/msi-laptop-bl/.
47 *
48 * This driver might work on other laptops produced by MSI. If you
49 * want to try it you can pass force=1 as argument to the module which
50 * will force it to load even when the DMI data doesn't identify the
51 * laptop as MSI S270. YMMV.
52 */
53
54#include <linux/module.h>
55#include <linux/kernel.h>
56#include <linux/init.h>
57#include <linux/acpi.h>
58#include <linux/dmi.h>
59#include <linux/backlight.h>
60#include <linux/platform_device.h>
61
62#define MSI_DRIVER_VERSION "0.5"
63
64#define MSI_LCD_LEVEL_MAX 9
65
66#define MSI_EC_COMMAND_WIRELESS 0x10
67#define MSI_EC_COMMAND_LCD_LEVEL 0x11
68
69static int force;
70module_param(force, bool, 0);
71MODULE_PARM_DESC(force, "Force driver load, ignore DMI data");
72
73static int auto_brightness;
74module_param(auto_brightness, int, 0);
75MODULE_PARM_DESC(auto_brightness, "Enable automatic brightness control (0: disabled; 1: enabled; 2: don't touch)");
76
77/* Hardware access */
78
79static int set_lcd_level(int level)
80{
81 u8 buf[2];
82
83 if (level < 0 || level >= MSI_LCD_LEVEL_MAX)
84 return -EINVAL;
85
86 buf[0] = 0x80;
87 buf[1] = (u8) (level*31);
88
89 return ec_transaction(MSI_EC_COMMAND_LCD_LEVEL, buf, sizeof(buf), NULL, 0, 1);
90}
91
92static int get_lcd_level(void)
93{
94 u8 wdata = 0, rdata;
95 int result;
96
97 result = ec_transaction(MSI_EC_COMMAND_LCD_LEVEL, &wdata, 1, &rdata, 1, 1);
98 if (result < 0)
99 return result;
100
101 return (int) rdata / 31;
102}
103
104static int get_auto_brightness(void)
105{
106 u8 wdata = 4, rdata;
107 int result;
108
109 result = ec_transaction(MSI_EC_COMMAND_LCD_LEVEL, &wdata, 1, &rdata, 1, 1);
110 if (result < 0)
111 return result;
112
113 return !!(rdata & 8);
114}
115
116static int set_auto_brightness(int enable)
117{
118 u8 wdata[2], rdata;
119 int result;
120
121 wdata[0] = 4;
122
123 result = ec_transaction(MSI_EC_COMMAND_LCD_LEVEL, wdata, 1, &rdata, 1, 1);
124 if (result < 0)
125 return result;
126
127 wdata[0] = 0x84;
128 wdata[1] = (rdata & 0xF7) | (enable ? 8 : 0);
129
130 return ec_transaction(MSI_EC_COMMAND_LCD_LEVEL, wdata, 2, NULL, 0, 1);
131}
132
133static int get_wireless_state(int *wlan, int *bluetooth)
134{
135 u8 wdata = 0, rdata;
136 int result;
137
138 result = ec_transaction(MSI_EC_COMMAND_WIRELESS, &wdata, 1, &rdata, 1, 1);
139 if (result < 0)
140 return -1;
141
142 if (wlan)
143 *wlan = !!(rdata & 8);
144
145 if (bluetooth)
146 *bluetooth = !!(rdata & 128);
147
148 return 0;
149}
150
151/* Backlight device stuff */
152
153static int bl_get_brightness(struct backlight_device *b)
154{
155 return get_lcd_level();
156}
157
158
159static int bl_update_status(struct backlight_device *b)
160{
161 return set_lcd_level(b->props.brightness);
162}
163
164static struct backlight_ops msibl_ops = {
165 .get_brightness = bl_get_brightness,
166 .update_status = bl_update_status,
167};
168
169static struct backlight_device *msibl_device;
170
171/* Platform device */
172
173static ssize_t show_wlan(struct device *dev,
174 struct device_attribute *attr, char *buf)
175{
176
177 int ret, enabled;
178
179 ret = get_wireless_state(&enabled, NULL);
180 if (ret < 0)
181 return ret;
182
183 return sprintf(buf, "%i\n", enabled);
184}
185
186static ssize_t show_bluetooth(struct device *dev,
187 struct device_attribute *attr, char *buf)
188{
189
190 int ret, enabled;
191
192 ret = get_wireless_state(NULL, &enabled);
193 if (ret < 0)
194 return ret;
195
196 return sprintf(buf, "%i\n", enabled);
197}
198
199static ssize_t show_lcd_level(struct device *dev,
200 struct device_attribute *attr, char *buf)
201{
202
203 int ret;
204
205 ret = get_lcd_level();
206 if (ret < 0)
207 return ret;
208
209 return sprintf(buf, "%i\n", ret);
210}
211
212static ssize_t store_lcd_level(struct device *dev,
213 struct device_attribute *attr, const char *buf, size_t count)
214{
215
216 int level, ret;
217
218 if (sscanf(buf, "%i", &level) != 1 || (level < 0 || level >= MSI_LCD_LEVEL_MAX))
219 return -EINVAL;
220
221 ret = set_lcd_level(level);
222 if (ret < 0)
223 return ret;
224
225 return count;
226}
227
228static ssize_t show_auto_brightness(struct device *dev,
229 struct device_attribute *attr, char *buf)
230{
231
232 int ret;
233
234 ret = get_auto_brightness();
235 if (ret < 0)
236 return ret;
237
238 return sprintf(buf, "%i\n", ret);
239}
240
241static ssize_t store_auto_brightness(struct device *dev,
242 struct device_attribute *attr, const char *buf, size_t count)
243{
244
245 int enable, ret;
246
247 if (sscanf(buf, "%i", &enable) != 1 || (enable != (enable & 1)))
248 return -EINVAL;
249
250 ret = set_auto_brightness(enable);
251 if (ret < 0)
252 return ret;
253
254 return count;
255}
256
257static DEVICE_ATTR(lcd_level, 0644, show_lcd_level, store_lcd_level);
258static DEVICE_ATTR(auto_brightness, 0644, show_auto_brightness, store_auto_brightness);
259static DEVICE_ATTR(bluetooth, 0444, show_bluetooth, NULL);
260static DEVICE_ATTR(wlan, 0444, show_wlan, NULL);
261
262static struct attribute *msipf_attributes[] = {
263 &dev_attr_lcd_level.attr,
264 &dev_attr_auto_brightness.attr,
265 &dev_attr_bluetooth.attr,
266 &dev_attr_wlan.attr,
267 NULL
268};
269
270static struct attribute_group msipf_attribute_group = {
271 .attrs = msipf_attributes
272};
273
274static struct platform_driver msipf_driver = {
275 .driver = {
276 .name = "msi-laptop-pf",
277 .owner = THIS_MODULE,
278 }
279};
280
281static struct platform_device *msipf_device;
282
283/* Initialization */
284
285static int dmi_check_cb(const struct dmi_system_id *id)
286{
287 printk("msi-laptop: Identified laptop model '%s'.\n", id->ident);
288 return 0;
289}
290
291static struct dmi_system_id __initdata msi_dmi_table[] = {
292 {
293 .ident = "MSI S270",
294 .matches = {
295 DMI_MATCH(DMI_SYS_VENDOR, "MICRO-STAR INT'L CO.,LTD"),
296 DMI_MATCH(DMI_PRODUCT_NAME, "MS-1013"),
297 DMI_MATCH(DMI_PRODUCT_VERSION, "0131"),
298 DMI_MATCH(DMI_CHASSIS_VENDOR, "MICRO-STAR INT'L CO.,LTD")
299 },
300 .callback = dmi_check_cb
301 },
302 {
303 .ident = "MSI S271",
304 .matches = {
305 DMI_MATCH(DMI_SYS_VENDOR, "Micro-Star International"),
306 DMI_MATCH(DMI_PRODUCT_NAME, "MS-1058"),
307 DMI_MATCH(DMI_PRODUCT_VERSION, "0581"),
308 DMI_MATCH(DMI_BOARD_NAME, "MS-1058")
309 },
310 .callback = dmi_check_cb
311 },
312 {
313 .ident = "MSI S420",
314 .matches = {
315 DMI_MATCH(DMI_SYS_VENDOR, "Micro-Star International"),
316 DMI_MATCH(DMI_PRODUCT_NAME, "MS-1412"),
317 DMI_MATCH(DMI_BOARD_VENDOR, "MSI"),
318 DMI_MATCH(DMI_BOARD_NAME, "MS-1412")
319 },
320 .callback = dmi_check_cb
321 },
322 {
323 .ident = "Medion MD96100",
324 .matches = {
325 DMI_MATCH(DMI_SYS_VENDOR, "NOTEBOOK"),
326 DMI_MATCH(DMI_PRODUCT_NAME, "SAM2000"),
327 DMI_MATCH(DMI_PRODUCT_VERSION, "0131"),
328 DMI_MATCH(DMI_CHASSIS_VENDOR, "MICRO-STAR INT'L CO.,LTD")
329 },
330 .callback = dmi_check_cb
331 },
332 { }
333};
334
335static int __init msi_init(void)
336{
337 int ret;
338
339 if (acpi_disabled)
340 return -ENODEV;
341
342 if (!force && !dmi_check_system(msi_dmi_table))
343 return -ENODEV;
344
345 if (auto_brightness < 0 || auto_brightness > 2)
346 return -EINVAL;
347
348 /* Register backlight stuff */
349
350 if (acpi_video_backlight_support()) {
351 printk(KERN_INFO "MSI: Brightness ignored, must be controlled "
352 "by ACPI video driver\n");
353 } else {
354 msibl_device = backlight_device_register("msi-laptop-bl", NULL,
355 NULL, &msibl_ops);
356 if (IS_ERR(msibl_device))
357 return PTR_ERR(msibl_device);
358 msibl_device->props.max_brightness = MSI_LCD_LEVEL_MAX-1;
359 }
360
361 ret = platform_driver_register(&msipf_driver);
362 if (ret)
363 goto fail_backlight;
364
365 /* Register platform stuff */
366
367 msipf_device = platform_device_alloc("msi-laptop-pf", -1);
368 if (!msipf_device) {
369 ret = -ENOMEM;
370 goto fail_platform_driver;
371 }
372
373 ret = platform_device_add(msipf_device);
374 if (ret)
375 goto fail_platform_device1;
376
377 ret = sysfs_create_group(&msipf_device->dev.kobj, &msipf_attribute_group);
378 if (ret)
379 goto fail_platform_device2;
380
381 /* Disable automatic brightness control by default because
382 * this module was probably loaded to do brightness control in
383 * software. */
384
385 if (auto_brightness != 2)
386 set_auto_brightness(auto_brightness);
387
388 printk(KERN_INFO "msi-laptop: driver "MSI_DRIVER_VERSION" successfully loaded.\n");
389
390 return 0;
391
392fail_platform_device2:
393
394 platform_device_del(msipf_device);
395
396fail_platform_device1:
397
398 platform_device_put(msipf_device);
399
400fail_platform_driver:
401
402 platform_driver_unregister(&msipf_driver);
403
404fail_backlight:
405
406 backlight_device_unregister(msibl_device);
407
408 return ret;
409}
410
411static void __exit msi_cleanup(void)
412{
413
414 sysfs_remove_group(&msipf_device->dev.kobj, &msipf_attribute_group);
415 platform_device_unregister(msipf_device);
416 platform_driver_unregister(&msipf_driver);
417 backlight_device_unregister(msibl_device);
418
419 /* Enable automatic brightness control again */
420 if (auto_brightness != 2)
421 set_auto_brightness(1);
422
423 printk(KERN_INFO "msi-laptop: driver unloaded.\n");
424}
425
426module_init(msi_init);
427module_exit(msi_cleanup);
428
429MODULE_AUTHOR("Lennart Poettering");
430MODULE_DESCRIPTION("MSI Laptop Support");
431MODULE_VERSION(MSI_DRIVER_VERSION);
432MODULE_LICENSE("GPL");
433
434MODULE_ALIAS("dmi:*:svnMICRO-STARINT'LCO.,LTD:pnMS-1013:pvr0131*:cvnMICRO-STARINT'LCO.,LTD:ct10:*");
435MODULE_ALIAS("dmi:*:svnMicro-StarInternational:pnMS-1058:pvr0581:rvnMSI:rnMS-1058:*:ct10:*");
436MODULE_ALIAS("dmi:*:svnMicro-StarInternational:pnMS-1412:*:rvnMSI:rnMS-1412:*:cvnMICRO-STARINT'LCO.,LTD:ct10:*");
437MODULE_ALIAS("dmi:*:svnNOTEBOOK:pnSAM2000:pvr0131*:cvnMICRO-STARINT'LCO.,LTD:ct10:*");
generated by cgit v1.2.2 (git 2.25.0) at 2024-12-30 12:13:42 -0500
>3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480 3481 3482 3483 3484 3485 3486 3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513 3514 3515 3516 3517 3518 3519 3520 3521 3522 3523 3524 3525 3526 3527 3528 3529 3530 3531 3532 3533 3534 3535 3536 3537 3538 3539 3540 3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551 3552 3553 3554 3555 3556 3557 3558 3559 3560 3561 3562 3563 3564 3565 3566 3567 3568 3569 3570 3571 3572 3573 3574 3575 3576 3577 3578 3579 3580 3581 3582 3583 3584 3585 3586 3587 3588 3589 3590 3591 3592 3593 3594 3595 3596 3597 3598 3599 3600 3601 3602 3603 3604 3605 3606 3607 3608 3609 3610 3611 3612 3613 3614 3615 3616 3617 3618 3619 3620 3621 3622 3623 3624 3625 3626 3627 3628 3629 3630 3631 3632 3633 3634 3635 3636 3637 3638 3639 3640 3641 3642 3643 3644 3645 3646 3647 3648 3649 3650 3651 3652 3653 3654 3655 3656 3657 3658 3659 3660 3661 3662 3663 3664 3665 3666 3667 3668 3669 3670 3671 3672 3673 3674 3675 3676 3677 3678 3679 3680 3681 3682 3683 3684 3685 3686 3687 3688 3689 3690 3691 3692 3693 3694 3695 3696 3697 3698 3699 3700 3701 3702 3703 3704 3705 3706 3707 3708 3709 3710 3711 3712 3713 3714 3715 3716 3717 3718 3719 3720 3721 3722 3723 3724 3725 3726 3727 3728 3729 3730 3731 3732 3733 3734 3735 3736 3737 3738 3739 3740 3741 3742 3743 3744 3745 3746 3747 3748 3749 3750 3751 3752 3753 3754 3755 3756 3757 3758 3759 3760 3761 3762 3763 3764 3765 3766 3767 3768 3769 3770 3771 3772 3773 3774 3775 3776 3777 3778 3779 3780 3781 3782 3783 3784 3785 3786 3787 3788 3789 3790 3791 3792 3793 3794 3795 3796 3797 3798 3799 3800 3801 3802 3803 3804 3805 3806 3807 3808 3809 3810 3811 3812 3813 3814 3815 3816 3817 3818 3819 3820 3821 3822 3823 3824 3825 3826 3827 3828 3829 3830 3831 3832 3833 3834 3835 3836 3837 3838 3839 3840 3841 3842 3843 3844 3845 3846 3847 3848 3849 3850 3851 3852 3853 3854 3855 3856 3857 3858 3859 3860 3861 3862 3863 3864 3865 3866 3867 3868 3869 3870 3871 3872 3873 3874 3875 3876 3877 3878 3879 3880 3881 3882 3883 3884 3885 3886 3887 3888 3889 3890 3891 3892 3893 3894 3895 3896 3897 3898 3899 3900 3901 3902 3903 3904 3905 3906 3907 3908 3909 3910 3911 3912 3913 3914 3915 3916 3917 3918 3919 3920 3921 3922 3923 3924 3925 3926 3927 3928 3929 3930 3931 3932 3933 3934 3935 3936 3937 3938 3939 3940 3941 3942 3943 3944 3945 3946 3947 3948 3949 3950 3951 3952 3953 3954 3955 3956 3957 3958 3959 3960 3961 3962 3963 3964 3965 3966 3967 3968 3969 3970 3971 3972 3973 3974 3975 3976 3977 3978 3979 3980 3981 3982 3983 3984 3985 3986 3987 3988 3989 3990 3991 3992 3993 3994 3995 3996 3997 3998 3999 4000 4001 4002 4003 4004 4005 4006 4007 4008 4009 4010 4011 4012 4013 4014 4015 4016 4017 4018 4019 4020 4021 4022 4023 4024 4025 4026 4027 4028 4029 4030 4031 4032 4033 4034 4035 4036 4037 4038 4039 4040 4041 4042 4043 4044 4045 4046 4047 4048 4049 4050 4051 4052 4053 4054 4055 4056 4057 4058 4059 4060 4061 4062 4063 4064 4065 4066 4067 4068 4069 4070 4071 4072 4073 4074 4075 4076 4077 4078 4079 4080 4081 4082 4083 4084 4085 4086 4087 4088 4089 4090 4091 4092 4093 4094 4095 4096 4097 4098 4099 4100 4101 4102 4103 4104 4105 4106 4107 4108 4109 4110 4111 4112 4113 4114 4115 4116 4117 4118 4119 4120 4121 4122 4123 4124 4125 4126 4127 4128 4129 4130 4131 4132 4133 4134 4135 4136 4137 4138 4139 4140 4141 4142 4143 4144 4145 4146 4147 4148 4149 4150 4151 4152 4153 4154 4155 4156 4157 4158 4159 4160 4161 4162 4163 4164 4165 4166 4167 4168 4169 4170 4171 4172 4173 4174 4175 4176 4177 4178 4179 4180 4181 4182 4183 4184 4185 4186 4187 4188 4189 4190 4191 4192 4193 4194 4195 4196 4197 4198 4199 4200 4201 4202 4203 4204 4205 4206 4207 4208 4209 4210 4211 4212 4213 4214 4215 4216 4217 4218 4219 4220 4221 4222 4223 4224 4225 4226 4227 4228 4229 4230 4231 4232 4233 4234 4235 4236 4237 4238 4239 4240 4241 4242 4243 4244 4245 4246 4247 4248 4249 4250 4251 4252 4253 4254 4255 4256 4257 4258 4259 4260 4261 4262 4263 4264 4265 4266 4267 4268 4269 4270 4271 4272 4273 4274 4275 4276 4277 4278 4279 4280 4281 4282 4283 4284 4285 4286 4287 4288 4289 4290 4291 4292 4293 4294 4295 4296 4297 4298 4299 4300 4301 4302 4303 4304 4305 4306 4307 4308 4309 4310 4311 4312 4313 4314 4315 4316 4317 4318 4319 4320 4321 4322 4323 4324 4325 4326 4327 4328 4329 4330 4331 4332 4333 4334 4335 4336 4337 4338 4339 4340 4341 4342 4343 4344 4345 4346 4347 4348 4349 4350 4351 4352 4353 4354 4355 4356 4357 4358 4359 4360 4361 4362 4363 4364 4365 4366 4367 4368 4369 4370 4371 4372 4373 4374 4375 4376 4377 4378 4379 4380 4381 4382 4383 4384 4385 4386 4387 4388 4389 4390 4391 4392 4393 4394 4395 4396 4397 4398 4399 4400 4401 4402 4403 4404 4405 4406 4407 4408 4409 4410 4411 4412 4413 4414 4415 4416 4417 4418 4419 4420 4421 4422 4423 4424 4425 4426 4427 4428 4429 4430 4431 4432 4433 4434 4435 4436 4437 4438 4439 4440 4441 4442 4443 4444 4445 4446 4447 4448 4449 4450 4451 4452 4453 4454 4455 4456 4457 4458 4459 4460 4461 4462 4463 4464 4465 4466 4467 4468 4469 4470 4471 4472 4473 4474 4475 4476 4477 4478 4479 4480 4481 4482 4483 4484 4485 4486 4487 4488 4489 4490 4491 4492 4493 4494 4495 4496 4497 4498 4499 4500 4501 4502 4503 4504 4505 4506 4507 4508 4509 4510 4511 4512 4513 4514 4515 4516 4517 4518 4519 4520 4521 4522 4523 4524 4525 4526 4527 4528 4529 4530 4531 4532 4533 4534 4535 4536 4537 4538 4539 4540 4541 4542 4543 4544 4545 4546 4547 4548 4549 4550 4551 4552 4553 4554 4555 4556 4557 4558 4559 4560 4561 4562 4563 4564 4565 4566 4567 4568 4569 4570 4571 4572 4573 4574 4575 4576 4577 4578 4579 4580 4581 4582 4583 4584 4585 4586 4587 4588 4589 4590 4591 4592 4593 4594 4595 4596 4597 4598 4599 4600 4601 4602 4603 4604 4605 4606 4607 4608 4609 4610 4611 4612 4613 4614 4615 4616 4617 4618 4619 4620 4621 4622 4623 4624 4625 4626 4627 4628 4629 4630 4631 4632 4633 4634 4635 4636 4637 4638 4639 4640 4641 4642 4643 4644 4645 4646 4647 4648 4649 4650 4651 4652 4653 4654 4655 4656 4657 4658 4659 4660 4661 4662 4663 4664 4665 4666 4667 4668 4669 4670 4671 4672 4673 4674 4675 4676 4677 4678 4679 4680 4681 4682 4683 4684 4685 4686 4687 4688 4689 4690 4691 4692 4693 4694 4695 4696 4697 4698 4699 4700 4701 4702 4703 4704 4705 4706 4707 4708 4709 4710 4711 4712 4713 4714 4715 4716 4717 4718 4719 4720 4721 4722 4723 4724 4725 4726 4727 4728 4729 4730 4731 4732 4733 4734 4735 4736 4737 4738 4739 4740 4741 4742 4743 4744 4745 4746 4747 4748 4749 4750 4751 4752 4753 4754 4755 4756 4757 4758 4759 4760 4761 4762 4763 4764 4765 4766 4767 4768 4769 4770 4771 4772 4773 4774 4775 4776 4777 4778 4779 4780 4781 4782 4783 4784 4785 4786 4787 4788 4789 4790 4791 4792 4793 4794 4795 4796 4797 4798 4799 4800 4801 4802 4803 4804 4805 4806 4807 4808 4809 4810 4811 4812 4813 4814 4815 4816 4817 4818 4819 4820 4821 4822 4823 4824 4825 4826 4827 4828 4829 4830 4831 4832 4833 4834 4835 4836 4837 4838 4839 4840 4841 4842 4843 4844 4845 4846 4847 4848 4849 4850 4851 4852 4853 4854 4855 4856 4857 4858 4859 4860 4861 4862 4863 4864 4865 4866 4867 4868 4869 4870 4871 4872 4873 4874 4875 4876 4877 4878 4879 4880 4881 4882 4883 4884 4885 4886 4887 4888 4889 4890 4891 4892 4893 4894 4895 4896 4897 4898 4899 4900 4901 4902 4903 4904 4905 4906 4907 4908 4909 4910 4911 4912 4913 4914 4915 4916 4917 4918 4919 4920 4921 4922 4923 4924 4925 4926 4927 4928 4929 4930 4931 4932 4933 4934 4935 4936 4937 4938 4939 4940 4941 4942 4943 4944 4945 4946 4947 4948 4949 4950 4951 4952 4953 4954 4955 4956 4957 4958 4959 4960 4961 4962 4963 4964 4965 4966 4967 4968 4969 4970 4971 4972 4973 4974 4975 4976 4977 4978 4979 4980 4981 4982 4983 4984 4985 4986 4987 4988 4989 4990 4991 4992 4993 4994 4995 4996 4997 4998 4999 5000 5001 5002 5003 5004 5005 5006 5007 5008 5009 5010 5011 5012 5013 5014 5015 5016 5017 5018 5019 5020 5021 5022 5023 5024 5025 5026 5027 5028 5029 5030 5031 5032 5033 5034 5035 5036 5037 5038 5039 5040 5041 5042 5043 5044 5045 5046 5047 5048 5049 5050 5051 5052 5053 5054 5055 5056 5057 5058 5059 5060 5061 5062 5063 5064 5065 5066 5067 5068 5069 5070 5071 5072 5073 5074 5075 5076 5077 5078 5079 5080 5081 5082 5083 5084 5085 5086 5087 5088 5089 5090 5091 5092 5093 5094 5095 5096 5097 5098 5099 5100 5101 5102 5103 5104 5105 5106 5107 5108 5109 5110 5111 5112 5113 5114 5115 5116 5117 5118 5119 5120 5121 5122 5123 5124 5125 5126 5127 5128 5129 5130 5131 5132 5133 5134 5135 5136 5137 5138 5139 5140 5141 5142 5143 5144 5145 5146 5147 5148 5149 5150 5151 5152 5153 5154 5155 5156 5157 5158 5159 5160 5161 5162 5163 5164 5165 5166 5167 5168 5169 5170 5171 5172 5173 5174 5175 5176 5177 5178 5179 5180 5181 5182 5183 5184 5185 5186 5187 5188 5189 5190 5191 5192 5193 5194 5195 5196 5197 5198 5199 5200 5201 5202 5203 5204 5205 5206 5207 5208 5209 5210 5211 5212 5213 5214 5215 5216 5217 5218 5219 5220 5221 5222 5223 5224 5225 5226 5227 5228 5229 5230 5231 5232 5233 5234 5235 5236 5237 5238 5239 5240 5241 5242 5243 5244 5245 5246 5247 5248 5249 5250 5251 5252 5253 5254 5255 5256 5257 5258 5259 5260 5261 5262 5263 5264 5265 5266 5267 5268 5269 5270 5271 5272 5273 5274 5275 5276 5277 5278 5279 5280 5281 5282 5283 5284 5285 5286 5287 5288 5289 5290 5291 5292 5293 5294 5295 5296 5297 5298 5299 5300 5301 5302 5303 5304 5305 5306 5307 5308 5309 5310 5311 
/*
 * Copyright (C) 2007 Oracle.  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 v2 as published by the Free Software Foundation.
 *
 * 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.  See the GNU
 * General Public License for more details.
 *
 * You should have received a copy of the GNU General Public
 * License along with this program; if not, write to the
 * Free Software Foundation, Inc., 59 Temple Place - Suite 330,
 * Boston, MA 021110-1307, USA.
 */

#include <linux/kernel.h>
#include <linux/bio.h>
#include <linux/buffer_head.h>
#include <linux/file.h>
#include <linux/fs.h>
#include <linux/pagemap.h>
#include <linux/highmem.h>
#include <linux/time.h>
#include <linux/init.h>
#include <linux/string.h>
#include <linux/backing-dev.h>
#include <linux/mpage.h>
#include <linux/swap.h>
#include <linux/writeback.h>
#include <linux/statfs.h>
#include <linux/compat.h>
#include <linux/bit_spinlock.h>
#include <linux/xattr.h>
#include <linux/posix_acl.h>
#include <linux/falloc.h>
#include "compat.h"
#include "ctree.h"
#include "disk-io.h"
#include "transaction.h"
#include "btrfs_inode.h"
#include "ioctl.h"
#include "print-tree.h"
#include "volumes.h"
#include "ordered-data.h"
#include "xattr.h"
#include "tree-log.h"
#include "compression.h"
#include "locking.h"

struct btrfs_iget_args {
	u64 ino;
	struct btrfs_root *root;
};

static const struct inode_operations btrfs_dir_inode_operations;
static const struct inode_operations btrfs_symlink_inode_operations;
static const struct inode_operations btrfs_dir_ro_inode_operations;
static const struct inode_operations btrfs_special_inode_operations;
static const struct inode_operations btrfs_file_inode_operations;
static const struct address_space_operations btrfs_aops;
static const struct address_space_operations btrfs_symlink_aops;
static struct file_operations btrfs_dir_file_operations;
static struct extent_io_ops btrfs_extent_io_ops;

static struct kmem_cache *btrfs_inode_cachep;
struct kmem_cache *btrfs_trans_handle_cachep;
struct kmem_cache *btrfs_transaction_cachep;
struct kmem_cache *btrfs_path_cachep;

#define S_SHIFT 12
static unsigned char btrfs_type_by_mode[S_IFMT >> S_SHIFT] = {
	[S_IFREG >> S_SHIFT]	= BTRFS_FT_REG_FILE,
	[S_IFDIR >> S_SHIFT]	= BTRFS_FT_DIR,
	[S_IFCHR >> S_SHIFT]	= BTRFS_FT_CHRDEV,
	[S_IFBLK >> S_SHIFT]	= BTRFS_FT_BLKDEV,
	[S_IFIFO >> S_SHIFT]	= BTRFS_FT_FIFO,
	[S_IFSOCK >> S_SHIFT]	= BTRFS_FT_SOCK,
	[S_IFLNK >> S_SHIFT]	= BTRFS_FT_SYMLINK,
};

static void btrfs_truncate(struct inode *inode);
static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end);
static noinline int cow_file_range(struct inode *inode,
				   struct page *locked_page,
				   u64 start, u64 end, int *page_started,
				   unsigned long *nr_written, int unlock);

static int btrfs_init_inode_security(struct inode *inode,  struct inode *dir)
{
	int err;

	err = btrfs_init_acl(inode, dir);
	if (!err)
		err = btrfs_xattr_security_init(inode, dir);
	return err;
}

/*
 * this does all the hard work for inserting an inline extent into
 * the btree.  The caller should have done a btrfs_drop_extents so that
 * no overlapping inline items exist in the btree
 */
static noinline int insert_inline_extent(struct btrfs_trans_handle *trans,
				struct btrfs_root *root, struct inode *inode,
				u64 start, size_t size, size_t compressed_size,
				struct page **compressed_pages)
{
	struct btrfs_key key;
	struct btrfs_path *path;
	struct extent_buffer *leaf;
	struct page *page = NULL;
	char *kaddr;
	unsigned long ptr;
	struct btrfs_file_extent_item *ei;
	int err = 0;
	int ret;
	size_t cur_size = size;
	size_t datasize;
	unsigned long offset;
	int use_compress = 0;

	if (compressed_size && compressed_pages) {
		use_compress = 1;
		cur_size = compressed_size;
	}

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	path->leave_spinning = 1;
	btrfs_set_trans_block_group(trans, inode);

	key.objectid = inode->i_ino;
	key.offset = start;
	btrfs_set_key_type(&key, BTRFS_EXTENT_DATA_KEY);
	datasize = btrfs_file_extent_calc_inline_size(cur_size);

	inode_add_bytes(inode, size);
	ret = btrfs_insert_empty_item(trans, root, path, &key,
				      datasize);
	BUG_ON(ret);
	if (ret) {
		err = ret;
		goto fail;
	}
	leaf = path->nodes[0];
	ei = btrfs_item_ptr(leaf, path->slots[0],
			    struct btrfs_file_extent_item);
	btrfs_set_file_extent_generation(leaf, ei, trans->transid);
	btrfs_set_file_extent_type(leaf, ei, BTRFS_FILE_EXTENT_INLINE);
	btrfs_set_file_extent_encryption(leaf, ei, 0);
	btrfs_set_file_extent_other_encoding(leaf, ei, 0);
	btrfs_set_file_extent_ram_bytes(leaf, ei, size);
	ptr = btrfs_file_extent_inline_start(ei);

	if (use_compress) {
		struct page *cpage;
		int i = 0;
		while (compressed_size > 0) {
			cpage = compressed_pages[i];
			cur_size = min_t(unsigned long, compressed_size,
				       PAGE_CACHE_SIZE);

			kaddr = kmap_atomic(cpage, KM_USER0);
			write_extent_buffer(leaf, kaddr, ptr, cur_size);
			kunmap_atomic(kaddr, KM_USER0);

			i++;
			ptr += cur_size;
			compressed_size -= cur_size;
		}
		btrfs_set_file_extent_compression(leaf, ei,
						  BTRFS_COMPRESS_ZLIB);
	} else {
		page = find_get_page(inode->i_mapping,
				     start >> PAGE_CACHE_SHIFT);
		btrfs_set_file_extent_compression(leaf, ei, 0);
		kaddr = kmap_atomic(page, KM_USER0);
		offset = start & (PAGE_CACHE_SIZE - 1);
		write_extent_buffer(leaf, kaddr + offset, ptr, size);
		kunmap_atomic(kaddr, KM_USER0);
		page_cache_release(page);
	}
	btrfs_mark_buffer_dirty(leaf);
	btrfs_free_path(path);

	BTRFS_I(inode)->disk_i_size = inode->i_size;
	btrfs_update_inode(trans, root, inode);
	return 0;
fail:
	btrfs_free_path(path);
	return err;
}


/*
 * conditionally insert an inline extent into the file.  This
 * does the checks required to make sure the data is small enough
 * to fit as an inline extent.
 */
static noinline int cow_file_range_inline(struct btrfs_trans_handle *trans,
				 struct btrfs_root *root,
				 struct inode *inode, u64 start, u64 end,
				 size_t compressed_size,
				 struct page **compressed_pages)
{
	u64 isize = i_size_read(inode);
	u64 actual_end = min(end + 1, isize);
	u64 inline_len = actual_end - start;
	u64 aligned_end = (end + root->sectorsize - 1) &
			~((u64)root->sectorsize - 1);
	u64 hint_byte;
	u64 data_len = inline_len;
	int ret;

	if (compressed_size)
		data_len = compressed_size;

	if (start > 0 ||
	    actual_end >= PAGE_CACHE_SIZE ||
	    data_len >= BTRFS_MAX_INLINE_DATA_SIZE(root) ||
	    (!compressed_size &&
	    (actual_end & (root->sectorsize - 1)) == 0) ||
	    end + 1 < isize ||
	    data_len > root->fs_info->max_inline) {
		return 1;
	}

	ret = btrfs_drop_extents(trans, root, inode, start,
				 aligned_end, aligned_end, start, &hint_byte);
	BUG_ON(ret);

	if (isize > actual_end)
		inline_len = min_t(u64, isize, actual_end);
	ret = insert_inline_extent(trans, root, inode, start,
				   inline_len, compressed_size,
				   compressed_pages);
	BUG_ON(ret);
	btrfs_drop_extent_cache(inode, start, aligned_end, 0);
	return 0;
}

struct async_extent {
	u64 start;
	u64 ram_size;
	u64 compressed_size;
	struct page **pages;
	unsigned long nr_pages;
	struct list_head list;
};

struct async_cow {
	struct inode *inode;
	struct btrfs_root *root;
	struct page *locked_page;
	u64 start;
	u64 end;
	struct list_head extents;
	struct btrfs_work work;
};

static noinline int add_async_extent(struct async_cow *cow,
				     u64 start, u64 ram_size,
				     u64 compressed_size,
				     struct page **pages,
				     unsigned long nr_pages)
{
	struct async_extent *async_extent;

	async_extent = kmalloc(sizeof(*async_extent), GFP_NOFS);
	async_extent->start = start;
	async_extent->ram_size = ram_size;
	async_extent->compressed_size = compressed_size;
	async_extent->pages = pages;
	async_extent->nr_pages = nr_pages;
	list_add_tail(&async_extent->list, &cow->extents);
	return 0;
}

/*
 * we create compressed extents in two phases.  The first
 * phase compresses a range of pages that have already been
 * locked (both pages and state bits are locked).
 *
 * This is done inside an ordered work queue, and the compression
 * is spread across many cpus.  The actual IO submission is step
 * two, and the ordered work queue takes care of making sure that
 * happens in the same order things were put onto the queue by
 * writepages and friends.
 *
 * If this code finds it can't get good compression, it puts an
 * entry onto the work queue to write the uncompressed bytes.  This
 * makes sure that both compressed inodes and uncompressed inodes
 * are written in the same order that pdflush sent them down.
 */
static noinline int compress_file_range(struct inode *inode,
					struct page *locked_page,
					u64 start, u64 end,
					struct async_cow *async_cow,
					int *num_added)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
	u64 num_bytes;
	u64 orig_start;
	u64 disk_num_bytes;
	u64 blocksize = root->sectorsize;
	u64 actual_end;
	u64 isize = i_size_read(inode);
	int ret = 0;
	struct page **pages = NULL;
	unsigned long nr_pages;
	unsigned long nr_pages_ret = 0;
	unsigned long total_compressed = 0;
	unsigned long total_in = 0;
	unsigned long max_compressed = 128 * 1024;
	unsigned long max_uncompressed = 128 * 1024;
	int i;
	int will_compress;

	orig_start = start;

	actual_end = min_t(u64, isize, end + 1);
again:
	will_compress = 0;
	nr_pages = (end >> PAGE_CACHE_SHIFT) - (start >> PAGE_CACHE_SHIFT) + 1;
	nr_pages = min(nr_pages, (128 * 1024UL) / PAGE_CACHE_SIZE);

	/*
	 * we don't want to send crud past the end of i_size through
	 * compression, that's just a waste of CPU time.  So, if the
	 * end of the file is before the start of our current
	 * requested range of bytes, we bail out to the uncompressed
	 * cleanup code that can deal with all of this.
	 *
	 * It isn't really the fastest way to fix things, but this is a
	 * very uncommon corner.
	 */
	if (actual_end <= start)
		goto cleanup_and_bail_uncompressed;

	total_compressed = actual_end - start;

	/* we want to make sure that amount of ram required to uncompress
	 * an extent is reasonable, so we limit the total size in ram
	 * of a compressed extent to 128k.  This is a crucial number
	 * because it also controls how easily we can spread reads across
	 * cpus for decompression.
	 *
	 * We also want to make sure the amount of IO required to do
	 * a random read is reasonably small, so we limit the size of
	 * a compressed extent to 128k.
	 */
	total_compressed = min(total_compressed, max_uncompressed);
	num_bytes = (end - start + blocksize) & ~(blocksize - 1);
	num_bytes = max(blocksize,  num_bytes);
	disk_num_bytes = num_bytes;
	total_in = 0;
	ret = 0;

	/*
	 * we do compression for mount -o compress and when the
	 * inode has not been flagged as nocompress.  This flag can
	 * change at any time if we discover bad compression ratios.
	 */
	if (!(BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS) &&
	    btrfs_test_opt(root, COMPRESS)) {
		WARN_ON(pages);
		pages = kzalloc(sizeof(struct page *) * nr_pages, GFP_NOFS);

		ret = btrfs_zlib_compress_pages(inode->i_mapping, start,
						total_compressed, pages,
						nr_pages, &nr_pages_ret,
						&total_in,
						&total_compressed,
						max_compressed);

		if (!ret) {
			unsigned long offset = total_compressed &
				(PAGE_CACHE_SIZE - 1);
			struct page *page = pages[nr_pages_ret - 1];
			char *kaddr;

			/* zero the tail end of the last page, we might be
			 * sending it down to disk
			 */
			if (offset) {
				kaddr = kmap_atomic(page, KM_USER0);
				memset(kaddr + offset, 0,
				       PAGE_CACHE_SIZE - offset);
				kunmap_atomic(kaddr, KM_USER0);
			}
			will_compress = 1;
		}
	}
	if (start == 0) {
		trans = btrfs_join_transaction(root, 1);
		BUG_ON(!trans);
		btrfs_set_trans_block_group(trans, inode);

		/* lets try to make an inline extent */
		if (ret || total_in < (actual_end - start)) {
			/* we didn't compress the entire range, try
			 * to make an uncompressed inline extent.
			 */
			ret = cow_file_range_inline(trans, root, inode,
						    start, end, 0, NULL);
		} else {
			/* try making a compressed inline extent */
			ret = cow_file_range_inline(trans, root, inode,
						    start, end,
						    total_compressed, pages);
		}
		btrfs_end_transaction(trans, root);
		if (ret == 0) {
			/*
			 * inline extent creation worked, we don't need
			 * to create any more async work items.  Unlock
			 * and free up our temp pages.
			 */
			extent_clear_unlock_delalloc(inode,
						     &BTRFS_I(inode)->io_tree,
						     start, end, NULL, 1, 0,
						     0, 1, 1, 1);
			ret = 0;
			goto free_pages_out;
		}
	}

	if (will_compress) {
		/*
		 * we aren't doing an inline extent round the compressed size
		 * up to a block size boundary so the allocator does sane
		 * things
		 */
		total_compressed = (total_compressed + blocksize - 1) &
			~(blocksize - 1);

		/*
		 * one last check to make sure the compression is really a
		 * win, compare the page count read with the blocks on disk
		 */
		total_in = (total_in + PAGE_CACHE_SIZE - 1) &
			~(PAGE_CACHE_SIZE - 1);
		if (total_compressed >= total_in) {
			will_compress = 0;
		} else {
			disk_num_bytes = total_compressed;
			num_bytes = total_in;
		}
	}
	if (!will_compress && pages) {
		/*
		 * the compression code ran but failed to make things smaller,
		 * free any pages it allocated and our page pointer array
		 */
		for (i = 0; i < nr_pages_ret; i++) {
			WARN_ON(pages[i]->mapping);
			page_cache_release(pages[i]);
		}
		kfree(pages);
		pages = NULL;
		total_compressed = 0;
		nr_pages_ret = 0;

		/* flag the file so we don't compress in the future */
		BTRFS_I(inode)->flags |= BTRFS_INODE_NOCOMPRESS;
	}
	if (will_compress) {
		*num_added += 1;

		/* the async work queues will take care of doing actual
		 * allocation on disk for these compressed pages,
		 * and will submit them to the elevator.
		 */
		add_async_extent(async_cow, start, num_bytes,
				 total_compressed, pages, nr_pages_ret);

		if (start + num_bytes < end && start + num_bytes < actual_end) {
			start += num_bytes;
			pages = NULL;
			cond_resched();
			goto again;
		}
	} else {
cleanup_and_bail_uncompressed:
		/*
		 * No compression, but we still need to write the pages in
		 * the file we've been given so far.  redirty the locked
		 * page if it corresponds to our extent and set things up
		 * for the async work queue to run cow_file_range to do
		 * the normal delalloc dance
		 */
		if (page_offset(locked_page) >= start &&
		    page_offset(locked_page) <= end) {
			__set_page_dirty_nobuffers(locked_page);
			/* unlocked later on in the async handlers */
		}
		add_async_extent(async_cow, start, end - start + 1, 0, NULL, 0);
		*num_added += 1;
	}

out:
	return 0;

free_pages_out:
	for (i = 0; i < nr_pages_ret; i++) {
		WARN_ON(pages[i]->mapping);
		page_cache_release(pages[i]);
	}
	kfree(pages);

	goto out;
}

/*
 * phase two of compressed writeback.  This is the ordered portion
 * of the code, which only gets called in the order the work was
 * queued.  We walk all the async extents created by compress_file_range
 * and send them down to the disk.
 */
static noinline int submit_compressed_extents(struct inode *inode,
					      struct async_cow *async_cow)
{
	struct async_extent *async_extent;
	u64 alloc_hint = 0;
	struct btrfs_trans_handle *trans;
	struct btrfs_key ins;
	struct extent_map *em;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
	struct extent_io_tree *io_tree;
	int ret;

	if (list_empty(&async_cow->extents))
		return 0;

	trans = btrfs_join_transaction(root, 1);

	while (!list_empty(&async_cow->extents)) {
		async_extent = list_entry(async_cow->extents.next,
					  struct async_extent, list);
		list_del(&async_extent->list);

		io_tree = &BTRFS_I(inode)->io_tree;

		/* did the compression code fall back to uncompressed IO? */
		if (!async_extent->pages) {
			int page_started = 0;
			unsigned long nr_written = 0;

			lock_extent(io_tree, async_extent->start,
				    async_extent->start +
				    async_extent->ram_size - 1, GFP_NOFS);

			/* allocate blocks */
			cow_file_range(inode, async_cow->locked_page,
				       async_extent->start,
				       async_extent->start +
				       async_extent->ram_size - 1,
				       &page_started, &nr_written, 0);

			/*
			 * if page_started, cow_file_range inserted an
			 * inline extent and took care of all the unlocking
			 * and IO for us.  Otherwise, we need to submit
			 * all those pages down to the drive.
			 */
			if (!page_started)
				extent_write_locked_range(io_tree,
						  inode, async_extent->start,
						  async_extent->start +
						  async_extent->ram_size - 1,
						  btrfs_get_extent,
						  WB_SYNC_ALL);
			kfree(async_extent);
			cond_resched();
			continue;
		}

		lock_extent(io_tree, async_extent->start,
			    async_extent->start + async_extent->ram_size - 1,
			    GFP_NOFS);
		/*
		 * here we're doing allocation and writeback of the
		 * compressed pages
		 */
		btrfs_drop_extent_cache(inode, async_extent->start,
					async_extent->start +
					async_extent->ram_size - 1, 0);

		ret = btrfs_reserve_extent(trans, root,
					   async_extent->compressed_size,
					   async_extent->compressed_size,
					   0, alloc_hint,
					   (u64)-1, &ins, 1);
		BUG_ON(ret);
		em = alloc_extent_map(GFP_NOFS);
		em->start = async_extent->start;
		em->len = async_extent->ram_size;
		em->orig_start = em->start;

		em->block_start = ins.objectid;
		em->block_len = ins.offset;
		em->bdev = root->fs_info->fs_devices->latest_bdev;
		set_bit(EXTENT_FLAG_PINNED, &em->flags);
		set_bit(EXTENT_FLAG_COMPRESSED, &em->flags);

		while (1) {
			spin_lock(&em_tree->lock);
			ret = add_extent_mapping(em_tree, em);
			spin_unlock(&em_tree->lock);
			if (ret != -EEXIST) {
				free_extent_map(em);
				break;
			}
			btrfs_drop_extent_cache(inode, async_extent->start,
						async_extent->start +
						async_extent->ram_size - 1, 0);
		}

		ret = btrfs_add_ordered_extent(inode, async_extent->start,
					       ins.objectid,
					       async_extent->ram_size,
					       ins.offset,
					       BTRFS_ORDERED_COMPRESSED);
		BUG_ON(ret);

		btrfs_end_transaction(trans, root);

		/*
		 * clear dirty, set writeback and unlock the pages.
		 */
		extent_clear_unlock_delalloc(inode,
					     &BTRFS_I(inode)->io_tree,
					     async_extent->start,
					     async_extent->start +
					     async_extent->ram_size - 1,
					     NULL, 1, 1, 0, 1, 1, 0);

		ret = btrfs_submit_compressed_write(inode,
				    async_extent->start,
				    async_extent->ram_size,
				    ins.objectid,
				    ins.offset, async_extent->pages,
				    async_extent->nr_pages);

		BUG_ON(ret);
		trans = btrfs_join_transaction(root, 1);
		alloc_hint = ins.objectid + ins.offset;
		kfree(async_extent);
		cond_resched();
	}

	btrfs_end_transaction(trans, root);
	return 0;
}

/*
 * when extent_io.c finds a delayed allocation range in the file,
 * the call backs end up in this code.  The basic idea is to
 * allocate extents on disk for the range, and create ordered data structs
 * in ram to track those extents.
 *
 * locked_page is the page that writepage had locked already.  We use
 * it to make sure we don't do extra locks or unlocks.
 *
 * *page_started is set to one if we unlock locked_page and do everything
 * required to start IO on it.  It may be clean and already done with
 * IO when we return.
 */
static noinline int cow_file_range(struct inode *inode,
				   struct page *locked_page,
				   u64 start, u64 end, int *page_started,
				   unsigned long *nr_written,
				   int unlock)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
	u64 alloc_hint = 0;
	u64 num_bytes;
	unsigned long ram_size;
	u64 disk_num_bytes;
	u64 cur_alloc_size;
	u64 blocksize = root->sectorsize;
	u64 actual_end;
	u64 isize = i_size_read(inode);
	struct btrfs_key ins;
	struct extent_map *em;
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
	int ret = 0;

	trans = btrfs_join_transaction(root, 1);
	BUG_ON(!trans);
	btrfs_set_trans_block_group(trans, inode);

	actual_end = min_t(u64, isize, end + 1);

	num_bytes = (end - start + blocksize) & ~(blocksize - 1);
	num_bytes = max(blocksize,  num_bytes);
	disk_num_bytes = num_bytes;
	ret = 0;

	if (start == 0) {
		/* lets try to make an inline extent */
		ret = cow_file_range_inline(trans, root, inode,
					    start, end, 0, NULL);
		if (ret == 0) {
			extent_clear_unlock_delalloc(inode,
						     &BTRFS_I(inode)->io_tree,
						     start, end, NULL, 1, 1,
						     1, 1, 1, 1);
			*nr_written = *nr_written +
			     (end - start + PAGE_CACHE_SIZE) / PAGE_CACHE_SIZE;
			*page_started = 1;
			ret = 0;
			goto out;
		}
	}

	BUG_ON(disk_num_bytes >
	       btrfs_super_total_bytes(&root->fs_info->super_copy));

	btrfs_drop_extent_cache(inode, start, start + num_bytes - 1, 0);

	while (disk_num_bytes > 0) {
		cur_alloc_size = min(disk_num_bytes, root->fs_info->max_extent);
		ret = btrfs_reserve_extent(trans, root, cur_alloc_size,
					   root->sectorsize, 0, alloc_hint,
					   (u64)-1, &ins, 1);
		BUG_ON(ret);

		em = alloc_extent_map(GFP_NOFS);
		em->start = start;
		em->orig_start = em->start;

		ram_size = ins.offset;
		em->len = ins.offset;

		em->block_start = ins.objectid;
		em->block_len = ins.offset;
		em->bdev = root->fs_info->fs_devices->latest_bdev;
		set_bit(EXTENT_FLAG_PINNED, &em->flags);

		while (1) {
			spin_lock(&em_tree->lock);
			ret = add_extent_mapping(em_tree, em);
			spin_unlock(&em_tree->lock);
			if (ret != -EEXIST) {
				free_extent_map(em);
				break;
			}
			btrfs_drop_extent_cache(inode, start,
						start + ram_size - 1, 0);
		}

		cur_alloc_size = ins.offset;
		ret = btrfs_add_ordered_extent(inode, start, ins.objectid,
					       ram_size, cur_alloc_size, 0);
		BUG_ON(ret);

		if (root->root_key.objectid ==
		    BTRFS_DATA_RELOC_TREE_OBJECTID) {
			ret = btrfs_reloc_clone_csums(inode, start,
						      cur_alloc_size);
			BUG_ON(ret);
		}

		if (disk_num_bytes < cur_alloc_size)
			break;

		/* we're not doing compressed IO, don't unlock the first
		 * page (which the caller expects to stay locked), don't
		 * clear any dirty bits and don't set any writeback bits
		 */
		extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree,
					     start, start + ram_size - 1,
					     locked_page, unlock, 1,
					     1, 0, 0, 0);
		disk_num_bytes -= cur_alloc_size;
		num_bytes -= cur_alloc_size;
		alloc_hint = ins.objectid + ins.offset;
		start += cur_alloc_size;
	}
out:
	ret = 0;
	btrfs_end_transaction(trans, root);

	return ret;
}

/*
 * work queue call back to started compression on a file and pages
 */
static noinline void async_cow_start(struct btrfs_work *work)
{
	struct async_cow *async_cow;
	int num_added = 0;
	async_cow = container_of(work, struct async_cow, work);

	compress_file_range(async_cow->inode, async_cow->locked_page,
			    async_cow->start, async_cow->end, async_cow,
			    &num_added);
	if (num_added == 0)
		async_cow->inode = NULL;
}

/*
 * work queue call back to submit previously compressed pages
 */
static noinline void async_cow_submit(struct btrfs_work *work)
{
	struct async_cow *async_cow;
	struct btrfs_root *root;
	unsigned long nr_pages;

	async_cow = container_of(work, struct async_cow, work);

	root = async_cow->root;
	nr_pages = (async_cow->end - async_cow->start + PAGE_CACHE_SIZE) >>
		PAGE_CACHE_SHIFT;

	atomic_sub(nr_pages, &root->fs_info->async_delalloc_pages);

	if (atomic_read(&root->fs_info->async_delalloc_pages) <
	    5 * 1042 * 1024 &&
	    waitqueue_active(&root->fs_info->async_submit_wait))
		wake_up(&root->fs_info->async_submit_wait);

	if (async_cow->inode)
		submit_compressed_extents(async_cow->inode, async_cow);
}

static noinline void async_cow_free(struct btrfs_work *work)
{
	struct async_cow *async_cow;
	async_cow = container_of(work, struct async_cow, work);
	kfree(async_cow);
}

static int cow_file_range_async(struct inode *inode, struct page *locked_page,
				u64 start, u64 end, int *page_started,
				unsigned long *nr_written)
{
	struct async_cow *async_cow;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	unsigned long nr_pages;
	u64 cur_end;
	int limit = 10 * 1024 * 1042;

	clear_extent_bit(&BTRFS_I(inode)->io_tree, start, end, EXTENT_LOCKED |
			 EXTENT_DELALLOC, 1, 0, GFP_NOFS);
	while (start < end) {
		async_cow = kmalloc(sizeof(*async_cow), GFP_NOFS);
		async_cow->inode = inode;
		async_cow->root = root;
		async_cow->locked_page = locked_page;
		async_cow->start = start;

		if (BTRFS_I(inode)->flags & BTRFS_INODE_NOCOMPRESS)
			cur_end = end;
		else
			cur_end = min(end, start + 512 * 1024 - 1);

		async_cow->end = cur_end;
		INIT_LIST_HEAD(&async_cow->extents);

		async_cow->work.func = async_cow_start;
		async_cow->work.ordered_func = async_cow_submit;
		async_cow->work.ordered_free = async_cow_free;
		async_cow->work.flags = 0;

		nr_pages = (cur_end - start + PAGE_CACHE_SIZE) >>
			PAGE_CACHE_SHIFT;
		atomic_add(nr_pages, &root->fs_info->async_delalloc_pages);

		btrfs_queue_worker(&root->fs_info->delalloc_workers,
				   &async_cow->work);

		if (atomic_read(&root->fs_info->async_delalloc_pages) > limit) {
			wait_event(root->fs_info->async_submit_wait,
			   (atomic_read(&root->fs_info->async_delalloc_pages) <
			    limit));
		}

		while (atomic_read(&root->fs_info->async_submit_draining) &&
		      atomic_read(&root->fs_info->async_delalloc_pages)) {
			wait_event(root->fs_info->async_submit_wait,
			  (atomic_read(&root->fs_info->async_delalloc_pages) ==
			   0));
		}

		*nr_written += nr_pages;
		start = cur_end + 1;
	}
	*page_started = 1;
	return 0;
}

static noinline int csum_exist_in_range(struct btrfs_root *root,
					u64 bytenr, u64 num_bytes)
{
	int ret;
	struct btrfs_ordered_sum *sums;
	LIST_HEAD(list);

	ret = btrfs_lookup_csums_range(root->fs_info->csum_root, bytenr,
				       bytenr + num_bytes - 1, &list);
	if (ret == 0 && list_empty(&list))
		return 0;

	while (!list_empty(&list)) {
		sums = list_entry(list.next, struct btrfs_ordered_sum, list);
		list_del(&sums->list);
		kfree(sums);
	}
	return 1;
}

/*
 * when nowcow writeback call back.  This checks for snapshots or COW copies
 * of the extents that exist in the file, and COWs the file as required.
 *
 * If no cow copies or snapshots exist, we write directly to the existing
 * blocks on disk
 */
static noinline int run_delalloc_nocow(struct inode *inode,
				       struct page *locked_page,
			      u64 start, u64 end, int *page_started, int force,
			      unsigned long *nr_written)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
	struct extent_buffer *leaf;
	struct btrfs_path *path;
	struct btrfs_file_extent_item *fi;
	struct btrfs_key found_key;
	u64 cow_start;
	u64 cur_offset;
	u64 extent_end;
	u64 extent_offset;
	u64 disk_bytenr;
	u64 num_bytes;
	int extent_type;
	int ret;
	int type;
	int nocow;
	int check_prev = 1;

	path = btrfs_alloc_path();
	BUG_ON(!path);
	trans = btrfs_join_transaction(root, 1);
	BUG_ON(!trans);

	cow_start = (u64)-1;
	cur_offset = start;
	while (1) {
		ret = btrfs_lookup_file_extent(trans, root, path, inode->i_ino,
					       cur_offset, 0);
		BUG_ON(ret < 0);
		if (ret > 0 && path->slots[0] > 0 && check_prev) {
			leaf = path->nodes[0];
			btrfs_item_key_to_cpu(leaf, &found_key,
					      path->slots[0] - 1);
			if (found_key.objectid == inode->i_ino &&
			    found_key.type == BTRFS_EXTENT_DATA_KEY)
				path->slots[0]--;
		}
		check_prev = 0;
next_slot:
		leaf = path->nodes[0];
		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
			ret = btrfs_next_leaf(root, path);
			if (ret < 0)
				BUG_ON(1);
			if (ret > 0)
				break;
			leaf = path->nodes[0];
		}

		nocow = 0;
		disk_bytenr = 0;
		num_bytes = 0;
		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);

		if (found_key.objectid > inode->i_ino ||
		    found_key.type > BTRFS_EXTENT_DATA_KEY ||
		    found_key.offset > end)
			break;

		if (found_key.offset > cur_offset) {
			extent_end = found_key.offset;
			goto out_check;
		}

		fi = btrfs_item_ptr(leaf, path->slots[0],
				    struct btrfs_file_extent_item);
		extent_type = btrfs_file_extent_type(leaf, fi);

		if (extent_type == BTRFS_FILE_EXTENT_REG ||
		    extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
			disk_bytenr = btrfs_file_extent_disk_bytenr(leaf, fi);
			extent_offset = btrfs_file_extent_offset(leaf, fi);
			extent_end = found_key.offset +
				btrfs_file_extent_num_bytes(leaf, fi);
			if (extent_end <= start) {
				path->slots[0]++;
				goto next_slot;
			}
			if (disk_bytenr == 0)
				goto out_check;
			if (btrfs_file_extent_compression(leaf, fi) ||
			    btrfs_file_extent_encryption(leaf, fi) ||
			    btrfs_file_extent_other_encoding(leaf, fi))
				goto out_check;
			if (extent_type == BTRFS_FILE_EXTENT_REG && !force)
				goto out_check;
			if (btrfs_extent_readonly(root, disk_bytenr))
				goto out_check;
			if (btrfs_cross_ref_exist(trans, root, inode->i_ino,
						  found_key.offset -
						  extent_offset, disk_bytenr))
				goto out_check;
			disk_bytenr += extent_offset;
			disk_bytenr += cur_offset - found_key.offset;
			num_bytes = min(end + 1, extent_end) - cur_offset;
			/*
			 * force cow if csum exists in the range.
			 * this ensure that csum for a given extent are
			 * either valid or do not exist.
			 */
			if (csum_exist_in_range(root, disk_bytenr, num_bytes))
				goto out_check;
			nocow = 1;
		} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
			extent_end = found_key.offset +
				btrfs_file_extent_inline_len(leaf, fi);
			extent_end = ALIGN(extent_end, root->sectorsize);
		} else {
			BUG_ON(1);
		}
out_check:
		if (extent_end <= start) {
			path->slots[0]++;
			goto next_slot;
		}
		if (!nocow) {
			if (cow_start == (u64)-1)
				cow_start = cur_offset;
			cur_offset = extent_end;
			if (cur_offset > end)
				break;
			path->slots[0]++;
			goto next_slot;
		}

		btrfs_release_path(root, path);
		if (cow_start != (u64)-1) {
			ret = cow_file_range(inode, locked_page, cow_start,
					found_key.offset - 1, page_started,
					nr_written, 1);
			BUG_ON(ret);
			cow_start = (u64)-1;
		}

		if (extent_type == BTRFS_FILE_EXTENT_PREALLOC) {
			struct extent_map *em;
			struct extent_map_tree *em_tree;
			em_tree = &BTRFS_I(inode)->extent_tree;
			em = alloc_extent_map(GFP_NOFS);
			em->start = cur_offset;
			em->orig_start = em->start;
			em->len = num_bytes;
			em->block_len = num_bytes;
			em->block_start = disk_bytenr;
			em->bdev = root->fs_info->fs_devices->latest_bdev;
			set_bit(EXTENT_FLAG_PINNED, &em->flags);
			while (1) {
				spin_lock(&em_tree->lock);
				ret = add_extent_mapping(em_tree, em);
				spin_unlock(&em_tree->lock);
				if (ret != -EEXIST) {
					free_extent_map(em);
					break;
				}
				btrfs_drop_extent_cache(inode, em->start,
						em->start + em->len - 1, 0);
			}
			type = BTRFS_ORDERED_PREALLOC;
		} else {
			type = BTRFS_ORDERED_NOCOW;
		}

		ret = btrfs_add_ordered_extent(inode, cur_offset, disk_bytenr,
					       num_bytes, num_bytes, type);
		BUG_ON(ret);

		extent_clear_unlock_delalloc(inode, &BTRFS_I(inode)->io_tree,
					cur_offset, cur_offset + num_bytes - 1,
					locked_page, 1, 1, 1, 0, 0, 0);
		cur_offset = extent_end;
		if (cur_offset > end)
			break;
	}
	btrfs_release_path(root, path);

	if (cur_offset <= end && cow_start == (u64)-1)
		cow_start = cur_offset;
	if (cow_start != (u64)-1) {
		ret = cow_file_range(inode, locked_page, cow_start, end,
				     page_started, nr_written, 1);
		BUG_ON(ret);
	}

	ret = btrfs_end_transaction(trans, root);
	BUG_ON(ret);
	btrfs_free_path(path);
	return 0;
}

/*
 * extent_io.c call back to do delayed allocation processing
 */
static int run_delalloc_range(struct inode *inode, struct page *locked_page,
			      u64 start, u64 end, int *page_started,
			      unsigned long *nr_written)
{
	int ret;
	struct btrfs_root *root = BTRFS_I(inode)->root;

	if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATACOW)
		ret = run_delalloc_nocow(inode, locked_page, start, end,
					 page_started, 1, nr_written);
	else if (BTRFS_I(inode)->flags & BTRFS_INODE_PREALLOC)
		ret = run_delalloc_nocow(inode, locked_page, start, end,
					 page_started, 0, nr_written);
	else if (!btrfs_test_opt(root, COMPRESS))
		ret = cow_file_range(inode, locked_page, start, end,
				      page_started, nr_written, 1);
	else
		ret = cow_file_range_async(inode, locked_page, start, end,
					   page_started, nr_written);
	return ret;
}

/*
 * extent_io.c set_bit_hook, used to track delayed allocation
 * bytes in this file, and to maintain the list of inodes that
 * have pending delalloc work to be done.
 */
static int btrfs_set_bit_hook(struct inode *inode, u64 start, u64 end,
		       unsigned long old, unsigned long bits)
{
	/*
	 * set_bit and clear bit hooks normally require _irqsave/restore
	 * but in this case, we are only testeing for the DELALLOC
	 * bit, which is only set or cleared with irqs on
	 */
	if (!(old & EXTENT_DELALLOC) && (bits & EXTENT_DELALLOC)) {
		struct btrfs_root *root = BTRFS_I(inode)->root;
		btrfs_delalloc_reserve_space(root, inode, end - start + 1);
		spin_lock(&root->fs_info->delalloc_lock);
		BTRFS_I(inode)->delalloc_bytes += end - start + 1;
		root->fs_info->delalloc_bytes += end - start + 1;
		if (list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
			list_add_tail(&BTRFS_I(inode)->delalloc_inodes,
				      &root->fs_info->delalloc_inodes);
		}
		spin_unlock(&root->fs_info->delalloc_lock);
	}
	return 0;
}

/*
 * extent_io.c clear_bit_hook, see set_bit_hook for why
 */
static int btrfs_clear_bit_hook(struct inode *inode, u64 start, u64 end,
			 unsigned long old, unsigned long bits)
{
	/*
	 * set_bit and clear bit hooks normally require _irqsave/restore
	 * but in this case, we are only testeing for the DELALLOC
	 * bit, which is only set or cleared with irqs on
	 */
	if ((old & EXTENT_DELALLOC) && (bits & EXTENT_DELALLOC)) {
		struct btrfs_root *root = BTRFS_I(inode)->root;

		spin_lock(&root->fs_info->delalloc_lock);
		if (end - start + 1 > root->fs_info->delalloc_bytes) {
			printk(KERN_INFO "btrfs warning: delalloc account "
			       "%llu %llu\n",
			       (unsigned long long)end - start + 1,
			       (unsigned long long)
			       root->fs_info->delalloc_bytes);
			btrfs_delalloc_free_space(root, inode, (u64)-1);
			root->fs_info->delalloc_bytes = 0;
			BTRFS_I(inode)->delalloc_bytes = 0;
		} else {
			btrfs_delalloc_free_space(root, inode,
						  end - start + 1);
			root->fs_info->delalloc_bytes -= end - start + 1;
			BTRFS_I(inode)->delalloc_bytes -= end - start + 1;
		}
		if (BTRFS_I(inode)->delalloc_bytes == 0 &&
		    !list_empty(&BTRFS_I(inode)->delalloc_inodes)) {
			list_del_init(&BTRFS_I(inode)->delalloc_inodes);
		}
		spin_unlock(&root->fs_info->delalloc_lock);
	}
	return 0;
}

/*
 * extent_io.c merge_bio_hook, this must check the chunk tree to make sure
 * we don't create bios that span stripes or chunks
 */
int btrfs_merge_bio_hook(struct page *page, unsigned long offset,
			 size_t size, struct bio *bio,
			 unsigned long bio_flags)
{
	struct btrfs_root *root = BTRFS_I(page->mapping->host)->root;
	struct btrfs_mapping_tree *map_tree;
	u64 logical = (u64)bio->bi_sector << 9;
	u64 length = 0;
	u64 map_length;
	int ret;

	if (bio_flags & EXTENT_BIO_COMPRESSED)
		return 0;

	length = bio->bi_size;
	map_tree = &root->fs_info->mapping_tree;
	map_length = length;
	ret = btrfs_map_block(map_tree, READ, logical,
			      &map_length, NULL, 0);

	if (map_length < length + size)
		return 1;
	return 0;
}

/*
 * in order to insert checksums into the metadata in large chunks,
 * we wait until bio submission time.   All the pages in the bio are
 * checksummed and sums are attached onto the ordered extent record.
 *
 * At IO completion time the cums attached on the ordered extent record
 * are inserted into the btree
 */
static int __btrfs_submit_bio_start(struct inode *inode, int rw,
				    struct bio *bio, int mirror_num,
				    unsigned long bio_flags)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret = 0;

	ret = btrfs_csum_one_bio(root, inode, bio, 0, 0);
	BUG_ON(ret);
	return 0;
}

/*
 * in order to insert checksums into the metadata in large chunks,
 * we wait until bio submission time.   All the pages in the bio are
 * checksummed and sums are attached onto the ordered extent record.
 *
 * At IO completion time the cums attached on the ordered extent record
 * are inserted into the btree
 */
static int __btrfs_submit_bio_done(struct inode *inode, int rw, struct bio *bio,
			  int mirror_num, unsigned long bio_flags)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	return btrfs_map_bio(root, rw, bio, mirror_num, 1);
}

/*
 * extent_io.c submission hook. This does the right thing for csum calculation
 * on write, or reading the csums from the tree before a read
 */
static int btrfs_submit_bio_hook(struct inode *inode, int rw, struct bio *bio,
			  int mirror_num, unsigned long bio_flags)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret = 0;
	int skip_sum;

	skip_sum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM;

	ret = btrfs_bio_wq_end_io(root->fs_info, bio, 0);
	BUG_ON(ret);

	if (!(rw & (1 << BIO_RW))) {
		if (bio_flags & EXTENT_BIO_COMPRESSED) {
			return btrfs_submit_compressed_read(inode, bio,
						    mirror_num, bio_flags);
		} else if (!skip_sum)
			btrfs_lookup_bio_sums(root, inode, bio, NULL);
		goto mapit;
	} else if (!skip_sum) {
		/* csum items have already been cloned */
		if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID)
			goto mapit;
		/* we're doing a write, do the async checksumming */
		return btrfs_wq_submit_bio(BTRFS_I(inode)->root->fs_info,
				   inode, rw, bio, mirror_num,
				   bio_flags, __btrfs_submit_bio_start,
				   __btrfs_submit_bio_done);
	}

mapit:
	return btrfs_map_bio(root, rw, bio, mirror_num, 0);
}

/*
 * given a list of ordered sums record them in the inode.  This happens
 * at IO completion time based on sums calculated at bio submission time.
 */
static noinline int add_pending_csums(struct btrfs_trans_handle *trans,
			     struct inode *inode, u64 file_offset,
			     struct list_head *list)
{
	struct btrfs_ordered_sum *sum;

	btrfs_set_trans_block_group(trans, inode);

	list_for_each_entry(sum, list, list) {
		btrfs_csum_file_blocks(trans,
		       BTRFS_I(inode)->root->fs_info->csum_root, sum);
	}
	return 0;
}

int btrfs_set_extent_delalloc(struct inode *inode, u64 start, u64 end)
{
	if ((end & (PAGE_CACHE_SIZE - 1)) == 0)
		WARN_ON(1);
	return set_extent_delalloc(&BTRFS_I(inode)->io_tree, start, end,
				   GFP_NOFS);
}

/* see btrfs_writepage_start_hook for details on why this is required */
struct btrfs_writepage_fixup {
	struct page *page;
	struct btrfs_work work;
};

static void btrfs_writepage_fixup_worker(struct btrfs_work *work)
{
	struct btrfs_writepage_fixup *fixup;
	struct btrfs_ordered_extent *ordered;
	struct page *page;
	struct inode *inode;
	u64 page_start;
	u64 page_end;

	fixup = container_of(work, struct btrfs_writepage_fixup, work);
	page = fixup->page;
again:
	lock_page(page);
	if (!page->mapping || !PageDirty(page) || !PageChecked(page)) {
		ClearPageChecked(page);
		goto out_page;
	}

	inode = page->mapping->host;
	page_start = page_offset(page);
	page_end = page_offset(page) + PAGE_CACHE_SIZE - 1;

	lock_extent(&BTRFS_I(inode)->io_tree, page_start, page_end, GFP_NOFS);

	/* already ordered? We're done */
	if (test_range_bit(&BTRFS_I(inode)->io_tree, page_start, page_end,
			     EXTENT_ORDERED, 0)) {
		goto out;
	}

	ordered = btrfs_lookup_ordered_extent(inode, page_start);
	if (ordered) {
		unlock_extent(&BTRFS_I(inode)->io_tree, page_start,
			      page_end, GFP_NOFS);
		unlock_page(page);
		btrfs_start_ordered_extent(inode, ordered, 1);
		goto again;
	}

	btrfs_set_extent_delalloc(inode, page_start, page_end);
	ClearPageChecked(page);
out:
	unlock_extent(&BTRFS_I(inode)->io_tree, page_start, page_end, GFP_NOFS);
out_page:
	unlock_page(page);
	page_cache_release(page);
}

/*
 * There are a few paths in the higher layers of the kernel that directly
 * set the page dirty bit without asking the filesystem if it is a
 * good idea.  This causes problems because we want to make sure COW
 * properly happens and the data=ordered rules are followed.
 *
 * In our case any range that doesn't have the ORDERED bit set
 * hasn't been properly setup for IO.  We kick off an async process
 * to fix it up.  The async helper will wait for ordered extents, set
 * the delalloc bit and make it safe to write the page.
 */
static int btrfs_writepage_start_hook(struct page *page, u64 start, u64 end)
{
	struct inode *inode = page->mapping->host;
	struct btrfs_writepage_fixup *fixup;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret;

	ret = test_range_bit(&BTRFS_I(inode)->io_tree, start, end,
			     EXTENT_ORDERED, 0);
	if (ret)
		return 0;

	if (PageChecked(page))
		return -EAGAIN;

	fixup = kzalloc(sizeof(*fixup), GFP_NOFS);
	if (!fixup)
		return -EAGAIN;

	SetPageChecked(page);
	page_cache_get(page);
	fixup->work.func = btrfs_writepage_fixup_worker;
	fixup->page = page;
	btrfs_queue_worker(&root->fs_info->fixup_workers, &fixup->work);
	return -EAGAIN;
}

static int insert_reserved_file_extent(struct btrfs_trans_handle *trans,
				       struct inode *inode, u64 file_pos,
				       u64 disk_bytenr, u64 disk_num_bytes,
				       u64 num_bytes, u64 ram_bytes,
				       u64 locked_end,
				       u8 compression, u8 encryption,
				       u16 other_encoding, int extent_type)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_file_extent_item *fi;
	struct btrfs_path *path;
	struct extent_buffer *leaf;
	struct btrfs_key ins;
	u64 hint;
	int ret;

	path = btrfs_alloc_path();
	BUG_ON(!path);

	path->leave_spinning = 1;
	ret = btrfs_drop_extents(trans, root, inode, file_pos,
				 file_pos + num_bytes, locked_end,
				 file_pos, &hint);
	BUG_ON(ret);

	ins.objectid = inode->i_ino;
	ins.offset = file_pos;
	ins.type = BTRFS_EXTENT_DATA_KEY;
	ret = btrfs_insert_empty_item(trans, root, path, &ins, sizeof(*fi));
	BUG_ON(ret);
	leaf = path->nodes[0];
	fi = btrfs_item_ptr(leaf, path->slots[0],
			    struct btrfs_file_extent_item);
	btrfs_set_file_extent_generation(leaf, fi, trans->transid);
	btrfs_set_file_extent_type(leaf, fi, extent_type);
	btrfs_set_file_extent_disk_bytenr(leaf, fi, disk_bytenr);
	btrfs_set_file_extent_disk_num_bytes(leaf, fi, disk_num_bytes);
	btrfs_set_file_extent_offset(leaf, fi, 0);
	btrfs_set_file_extent_num_bytes(leaf, fi, num_bytes);
	btrfs_set_file_extent_ram_bytes(leaf, fi, ram_bytes);
	btrfs_set_file_extent_compression(leaf, fi, compression);
	btrfs_set_file_extent_encryption(leaf, fi, encryption);
	btrfs_set_file_extent_other_encoding(leaf, fi, other_encoding);

	btrfs_unlock_up_safe(path, 1);
	btrfs_set_lock_blocking(leaf);

	btrfs_mark_buffer_dirty(leaf);

	inode_add_bytes(inode, num_bytes);
	btrfs_drop_extent_cache(inode, file_pos, file_pos + num_bytes - 1, 0);

	ins.objectid = disk_bytenr;
	ins.offset = disk_num_bytes;
	ins.type = BTRFS_EXTENT_ITEM_KEY;
	ret = btrfs_alloc_reserved_file_extent(trans, root,
					root->root_key.objectid,
					inode->i_ino, file_pos, &ins);
	BUG_ON(ret);
	btrfs_free_path(path);

	return 0;
}

/*
 * helper function for btrfs_finish_ordered_io, this
 * just reads in some of the csum leaves to prime them into ram
 * before we start the transaction.  It limits the amount of btree
 * reads required while inside the transaction.
 */
static noinline void reada_csum(struct btrfs_root *root,
				struct btrfs_path *path,
				struct btrfs_ordered_extent *ordered_extent)
{
	struct btrfs_ordered_sum *sum;
	u64 bytenr;

	sum = list_entry(ordered_extent->list.next, struct btrfs_ordered_sum,
			 list);
	bytenr = sum->sums[0].bytenr;

	/*
	 * we don't care about the results, the point of this search is
	 * just to get the btree leaves into ram
	 */
	btrfs_lookup_csum(NULL, root->fs_info->csum_root, path, bytenr, 0);
}

/* as ordered data IO finishes, this gets called so we can finish
 * an ordered extent if the range of bytes in the file it covers are
 * fully written.
 */
static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
	struct btrfs_ordered_extent *ordered_extent = NULL;
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct btrfs_path *path;
	int compressed = 0;
	int ret;

	ret = btrfs_dec_test_ordered_pending(inode, start, end - start + 1);
	if (!ret)
		return 0;

	/*
	 * before we join the transaction, try to do some of our IO.
	 * This will limit the amount of IO that we have to do with
	 * the transaction running.  We're unlikely to need to do any
	 * IO if the file extents are new, the disk_i_size checks
	 * covers the most common case.
	 */
	if (start < BTRFS_I(inode)->disk_i_size) {
		path = btrfs_alloc_path();
		if (path) {
			ret = btrfs_lookup_file_extent(NULL, root, path,
						       inode->i_ino,
						       start, 0);
			ordered_extent = btrfs_lookup_ordered_extent(inode,
								     start);
			if (!list_empty(&ordered_extent->list)) {
				btrfs_release_path(root, path);
				reada_csum(root, path, ordered_extent);
			}
			btrfs_free_path(path);
		}
	}

	trans = btrfs_join_transaction(root, 1);

	if (!ordered_extent)
		ordered_extent = btrfs_lookup_ordered_extent(inode, start);
	BUG_ON(!ordered_extent);
	if (test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags))
		goto nocow;

	lock_extent(io_tree, ordered_extent->file_offset,
		    ordered_extent->file_offset + ordered_extent->len - 1,
		    GFP_NOFS);

	if (test_bit(BTRFS_ORDERED_COMPRESSED, &ordered_extent->flags))
		compressed = 1;
	if (test_bit(BTRFS_ORDERED_PREALLOC, &ordered_extent->flags)) {
		BUG_ON(compressed);
		ret = btrfs_mark_extent_written(trans, root, inode,
						ordered_extent->file_offset,
						ordered_extent->file_offset +
						ordered_extent->len);
		BUG_ON(ret);
	} else {
		ret = insert_reserved_file_extent(trans, inode,
						ordered_extent->file_offset,
						ordered_extent->start,
						ordered_extent->disk_len,
						ordered_extent->len,
						ordered_extent->len,
						ordered_extent->file_offset +
						ordered_extent->len,
						compressed, 0, 0,
						BTRFS_FILE_EXTENT_REG);
		BUG_ON(ret);
	}
	unlock_extent(io_tree, ordered_extent->file_offset,
		    ordered_extent->file_offset + ordered_extent->len - 1,
		    GFP_NOFS);
nocow:
	add_pending_csums(trans, inode, ordered_extent->file_offset,
			  &ordered_extent->list);

	mutex_lock(&BTRFS_I(inode)->extent_mutex);
	btrfs_ordered_update_i_size(inode, ordered_extent);
	btrfs_update_inode(trans, root, inode);
	btrfs_remove_ordered_extent(inode, ordered_extent);
	mutex_unlock(&BTRFS_I(inode)->extent_mutex);

	/* once for us */
	btrfs_put_ordered_extent(ordered_extent);
	/* once for the tree */
	btrfs_put_ordered_extent(ordered_extent);

	btrfs_end_transaction(trans, root);
	return 0;
}

static int btrfs_writepage_end_io_hook(struct page *page, u64 start, u64 end,
				struct extent_state *state, int uptodate)
{
	return btrfs_finish_ordered_io(page->mapping->host, start, end);
}

/*
 * When IO fails, either with EIO or csum verification fails, we
 * try other mirrors that might have a good copy of the data.  This
 * io_failure_record is used to record state as we go through all the
 * mirrors.  If another mirror has good data, the page is set up to date
 * and things continue.  If a good mirror can't be found, the original
 * bio end_io callback is called to indicate things have failed.
 */
struct io_failure_record {
	struct page *page;
	u64 start;
	u64 len;
	u64 logical;
	unsigned long bio_flags;
	int last_mirror;
};

static int btrfs_io_failed_hook(struct bio *failed_bio,
			 struct page *page, u64 start, u64 end,
			 struct extent_state *state)
{
	struct io_failure_record *failrec = NULL;
	u64 private;
	struct extent_map *em;
	struct inode *inode = page->mapping->host;
	struct extent_io_tree *failure_tree = &BTRFS_I(inode)->io_failure_tree;
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
	struct bio *bio;
	int num_copies;
	int ret;
	int rw;
	u64 logical;

	ret = get_state_private(failure_tree, start, &private);
	if (ret) {
		failrec = kmalloc(sizeof(*failrec), GFP_NOFS);
		if (!failrec)
			return -ENOMEM;
		failrec->start = start;
		failrec->len = end - start + 1;
		failrec->last_mirror = 0;
		failrec->bio_flags = 0;

		spin_lock(&em_tree->lock);
		em = lookup_extent_mapping(em_tree, start, failrec->len);
		if (em->start > start || em->start + em->len < start) {
			free_extent_map(em);
			em = NULL;
		}
		spin_unlock(&em_tree->lock);

		if (!em || IS_ERR(em)) {
			kfree(failrec);
			return -EIO;
		}
		logical = start - em->start;
		logical = em->block_start + logical;
		if (test_bit(EXTENT_FLAG_COMPRESSED, &em->flags)) {
			logical = em->block_start;
			failrec->bio_flags = EXTENT_BIO_COMPRESSED;
		}
		failrec->logical = logical;
		free_extent_map(em);
		set_extent_bits(failure_tree, start, end, EXTENT_LOCKED |
				EXTENT_DIRTY, GFP_NOFS);
		set_state_private(failure_tree, start,
				 (u64)(unsigned long)failrec);
	} else {
		failrec = (struct io_failure_record *)(unsigned long)private;
	}
	num_copies = btrfs_num_copies(
			      &BTRFS_I(inode)->root->fs_info->mapping_tree,
			      failrec->logical, failrec->len);
	failrec->last_mirror++;
	if (!state) {
		spin_lock(&BTRFS_I(inode)->io_tree.lock);
		state = find_first_extent_bit_state(&BTRFS_I(inode)->io_tree,
						    failrec->start,
						    EXTENT_LOCKED);
		if (state && state->start != failrec->start)
			state = NULL;
		spin_unlock(&BTRFS_I(inode)->io_tree.lock);
	}
	if (!state || failrec->last_mirror > num_copies) {
		set_state_private(failure_tree, failrec->start, 0);
		clear_extent_bits(failure_tree, failrec->start,
				  failrec->start + failrec->len - 1,
				  EXTENT_LOCKED | EXTENT_DIRTY, GFP_NOFS);
		kfree(failrec);
		return -EIO;
	}
	bio = bio_alloc(GFP_NOFS, 1);
	bio->bi_private = state;
	bio->bi_end_io = failed_bio->bi_end_io;
	bio->bi_sector = failrec->logical >> 9;
	bio->bi_bdev = failed_bio->bi_bdev;
	bio->bi_size = 0;

	bio_add_page(bio, page, failrec->len, start - page_offset(page));
	if (failed_bio->bi_rw & (1 << BIO_RW))
		rw = WRITE;
	else
		rw = READ;

	BTRFS_I(inode)->io_tree.ops->submit_bio_hook(inode, rw, bio,
						      failrec->last_mirror,
						      failrec->bio_flags);
	return 0;
}

/*
 * each time an IO finishes, we do a fast check in the IO failure tree
 * to see if we need to process or clean up an io_failure_record
 */
static int btrfs_clean_io_failures(struct inode *inode, u64 start)
{
	u64 private;
	u64 private_failure;
	struct io_failure_record *failure;
	int ret;

	private = 0;
	if (count_range_bits(&BTRFS_I(inode)->io_failure_tree, &private,
			     (u64)-1, 1, EXTENT_DIRTY)) {
		ret = get_state_private(&BTRFS_I(inode)->io_failure_tree,
					start, &private_failure);
		if (ret == 0) {
			failure = (struct io_failure_record *)(unsigned long)
				   private_failure;
			set_state_private(&BTRFS_I(inode)->io_failure_tree,
					  failure->start, 0);
			clear_extent_bits(&BTRFS_I(inode)->io_failure_tree,
					  failure->start,
					  failure->start + failure->len - 1,
					  EXTENT_DIRTY | EXTENT_LOCKED,
					  GFP_NOFS);
			kfree(failure);
		}
	}
	return 0;
}

/*
 * when reads are done, we need to check csums to verify the data is correct
 * if there's a match, we allow the bio to finish.  If not, we go through
 * the io_failure_record routines to find good copies
 */
static int btrfs_readpage_end_io_hook(struct page *page, u64 start, u64 end,
			       struct extent_state *state)
{
	size_t offset = start - ((u64)page->index << PAGE_CACHE_SHIFT);
	struct inode *inode = page->mapping->host;
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	char *kaddr;
	u64 private = ~(u32)0;
	int ret;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	u32 csum = ~(u32)0;

	if (PageChecked(page)) {
		ClearPageChecked(page);
		goto good;
	}

	if (BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM)
		return 0;

	if (root->root_key.objectid == BTRFS_DATA_RELOC_TREE_OBJECTID &&
	    test_range_bit(io_tree, start, end, EXTENT_NODATASUM, 1)) {
		clear_extent_bits(io_tree, start, end, EXTENT_NODATASUM,
				  GFP_NOFS);
		return 0;
	}

	if (state && state->start == start) {
		private = state->private;
		ret = 0;
	} else {
		ret = get_state_private(io_tree, start, &private);
	}
	kaddr = kmap_atomic(page, KM_USER0);
	if (ret)
		goto zeroit;

	csum = btrfs_csum_data(root, kaddr + offset, csum,  end - start + 1);
	btrfs_csum_final(csum, (char *)&csum);
	if (csum != private)
		goto zeroit;

	kunmap_atomic(kaddr, KM_USER0);
good:
	/* if the io failure tree for this inode is non-empty,
	 * check to see if we've recovered from a failed IO
	 */
	btrfs_clean_io_failures(inode, start);
	return 0;

zeroit:
	if (printk_ratelimit()) {
		printk(KERN_INFO "btrfs csum failed ino %lu off %llu csum %u "
		       "private %llu\n", page->mapping->host->i_ino,
		       (unsigned long long)start, csum,
		       (unsigned long long)private);
	}
	memset(kaddr + offset, 1, end - start + 1);
	flush_dcache_page(page);
	kunmap_atomic(kaddr, KM_USER0);
	if (private == 0)
		return 0;
	return -EIO;
}

/*
 * This creates an orphan entry for the given inode in case something goes
 * wrong in the middle of an unlink/truncate.
 */
int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret = 0;

	spin_lock(&root->list_lock);

	/* already on the orphan list, we're good */
	if (!list_empty(&BTRFS_I(inode)->i_orphan)) {
		spin_unlock(&root->list_lock);
		return 0;
	}

	list_add(&BTRFS_I(inode)->i_orphan, &root->orphan_list);

	spin_unlock(&root->list_lock);

	/*
	 * insert an orphan item to track this unlinked/truncated file
	 */
	ret = btrfs_insert_orphan_item(trans, root, inode->i_ino);

	return ret;
}

/*
 * We have done the truncate/delete so we can go ahead and remove the orphan
 * item for this particular inode.
 */
int btrfs_orphan_del(struct btrfs_trans_handle *trans, struct inode *inode)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret = 0;

	spin_lock(&root->list_lock);

	if (list_empty(&BTRFS_I(inode)->i_orphan)) {
		spin_unlock(&root->list_lock);
		return 0;
	}

	list_del_init(&BTRFS_I(inode)->i_orphan);
	if (!trans) {
		spin_unlock(&root->list_lock);
		return 0;
	}

	spin_unlock(&root->list_lock);

	ret = btrfs_del_orphan_item(trans, root, inode->i_ino);

	return ret;
}

/*
 * this cleans up any orphans that may be left on the list from the last use
 * of this root.
 */
void btrfs_orphan_cleanup(struct btrfs_root *root)
{
	struct btrfs_path *path;
	struct extent_buffer *leaf;
	struct btrfs_item *item;
	struct btrfs_key key, found_key;
	struct btrfs_trans_handle *trans;
	struct inode *inode;
	int ret = 0, nr_unlink = 0, nr_truncate = 0;

	path = btrfs_alloc_path();
	if (!path)
		return;
	path->reada = -1;

	key.objectid = BTRFS_ORPHAN_OBJECTID;
	btrfs_set_key_type(&key, BTRFS_ORPHAN_ITEM_KEY);
	key.offset = (u64)-1;


	while (1) {
		ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
		if (ret < 0) {
			printk(KERN_ERR "Error searching slot for orphan: %d"
			       "\n", ret);
			break;
		}

		/*
		 * if ret == 0 means we found what we were searching for, which
		 * is weird, but possible, so only screw with path if we didnt
		 * find the key and see if we have stuff that matches
		 */
		if (ret > 0) {
			if (path->slots[0] == 0)
				break;
			path->slots[0]--;
		}

		/* pull out the item */
		leaf = path->nodes[0];
		item = btrfs_item_nr(leaf, path->slots[0]);
		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);

		/* make sure the item matches what we want */
		if (found_key.objectid != BTRFS_ORPHAN_OBJECTID)
			break;
		if (btrfs_key_type(&found_key) != BTRFS_ORPHAN_ITEM_KEY)
			break;

		/* release the path since we're done with it */
		btrfs_release_path(root, path);

		/*
		 * this is where we are basically btrfs_lookup, without the
		 * crossing root thing.  we store the inode number in the
		 * offset of the orphan item.
		 */
		found_key.objectid = found_key.offset;
		found_key.type = BTRFS_INODE_ITEM_KEY;
		found_key.offset = 0;
		inode = btrfs_iget(root->fs_info->sb, &found_key, root);
		if (IS_ERR(inode))
			break;

		/*
		 * add this inode to the orphan list so btrfs_orphan_del does
		 * the proper thing when we hit it
		 */
		spin_lock(&root->list_lock);
		list_add(&BTRFS_I(inode)->i_orphan, &root->orphan_list);
		spin_unlock(&root->list_lock);

		/*
		 * if this is a bad inode, means we actually succeeded in
		 * removing the inode, but not the orphan record, which means
		 * we need to manually delete the orphan since iput will just
		 * do a destroy_inode
		 */
		if (is_bad_inode(inode)) {
			trans = btrfs_start_transaction(root, 1);
			btrfs_orphan_del(trans, inode);
			btrfs_end_transaction(trans, root);
			iput(inode);
			continue;
		}

		/* if we have links, this was a truncate, lets do that */
		if (inode->i_nlink) {
			nr_truncate++;
			btrfs_truncate(inode);
		} else {
			nr_unlink++;
		}

		/* this will do delete_inode and everything for us */
		iput(inode);
	}

	if (nr_unlink)
		printk(KERN_INFO "btrfs: unlinked %d orphans\n", nr_unlink);
	if (nr_truncate)
		printk(KERN_INFO "btrfs: truncated %d orphans\n", nr_truncate);

	btrfs_free_path(path);
}

/*
 * very simple check to peek ahead in the leaf looking for xattrs.  If we
 * don't find any xattrs, we know there can't be any acls.
 *
 * slot is the slot the inode is in, objectid is the objectid of the inode
 */
static noinline int acls_after_inode_item(struct extent_buffer *leaf,
					  int slot, u64 objectid)
{
	u32 nritems = btrfs_header_nritems(leaf);
	struct btrfs_key found_key;
	int scanned = 0;

	slot++;
	while (slot < nritems) {
		btrfs_item_key_to_cpu(leaf, &found_key, slot);

		/* we found a different objectid, there must not be acls */
		if (found_key.objectid != objectid)
			return 0;

		/* we found an xattr, assume we've got an acl */
		if (found_key.type == BTRFS_XATTR_ITEM_KEY)
			return 1;

		/*
		 * we found a key greater than an xattr key, there can't
		 * be any acls later on
		 */
		if (found_key.type > BTRFS_XATTR_ITEM_KEY)
			return 0;

		slot++;
		scanned++;

		/*
		 * it goes inode, inode backrefs, xattrs, extents,
		 * so if there are a ton of hard links to an inode there can
		 * be a lot of backrefs.  Don't waste time searching too hard,
		 * this is just an optimization
		 */
		if (scanned >= 8)
			break;
	}
	/* we hit the end of the leaf before we found an xattr or
	 * something larger than an xattr.  We have to assume the inode
	 * has acls
	 */
	return 1;
}

/*
 * read an inode from the btree into the in-memory inode
 */
static void btrfs_read_locked_inode(struct inode *inode)
{
	struct btrfs_path *path;
	struct extent_buffer *leaf;
	struct btrfs_inode_item *inode_item;
	struct btrfs_timespec *tspec;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_key location;
	int maybe_acls;
	u64 alloc_group_block;
	u32 rdev;
	int ret;

	path = btrfs_alloc_path();
	BUG_ON(!path);
	memcpy(&location, &BTRFS_I(inode)->location, sizeof(location));

	ret = btrfs_lookup_inode(NULL, root, path, &location, 0);
	if (ret)
		goto make_bad;

	leaf = path->nodes[0];
	inode_item = btrfs_item_ptr(leaf, path->slots[0],
				    struct btrfs_inode_item);

	inode->i_mode = btrfs_inode_mode(leaf, inode_item);
	inode->i_nlink = btrfs_inode_nlink(leaf, inode_item);
	inode->i_uid = btrfs_inode_uid(leaf, inode_item);
	inode->i_gid = btrfs_inode_gid(leaf, inode_item);
	btrfs_i_size_write(inode, btrfs_inode_size(leaf, inode_item));

	tspec = btrfs_inode_atime(inode_item);
	inode->i_atime.tv_sec = btrfs_timespec_sec(leaf, tspec);
	inode->i_atime.tv_nsec = btrfs_timespec_nsec(leaf, tspec);

	tspec = btrfs_inode_mtime(inode_item);
	inode->i_mtime.tv_sec = btrfs_timespec_sec(leaf, tspec);
	inode->i_mtime.tv_nsec = btrfs_timespec_nsec(leaf, tspec);

	tspec = btrfs_inode_ctime(inode_item);
	inode->i_ctime.tv_sec = btrfs_timespec_sec(leaf, tspec);
	inode->i_ctime.tv_nsec = btrfs_timespec_nsec(leaf, tspec);

	inode_set_bytes(inode, btrfs_inode_nbytes(leaf, inode_item));
	BTRFS_I(inode)->generation = btrfs_inode_generation(leaf, inode_item);
	BTRFS_I(inode)->sequence = btrfs_inode_sequence(leaf, inode_item);
	inode->i_generation = BTRFS_I(inode)->generation;
	inode->i_rdev = 0;
	rdev = btrfs_inode_rdev(leaf, inode_item);

	BTRFS_I(inode)->index_cnt = (u64)-1;
	BTRFS_I(inode)->flags = btrfs_inode_flags(leaf, inode_item);

	alloc_group_block = btrfs_inode_block_group(leaf, inode_item);

	/*
	 * try to precache a NULL acl entry for files that don't have
	 * any xattrs or acls
	 */
	maybe_acls = acls_after_inode_item(leaf, path->slots[0], inode->i_ino);
	if (!maybe_acls)
		cache_no_acl(inode);

	BTRFS_I(inode)->block_group = btrfs_find_block_group(root, 0,
						alloc_group_block, 0);
	btrfs_free_path(path);
	inode_item = NULL;

	switch (inode->i_mode & S_IFMT) {
	case S_IFREG:
		inode->i_mapping->a_ops = &btrfs_aops;
		inode->i_mapping->backing_dev_info = &root->fs_info->bdi;
		BTRFS_I(inode)->io_tree.ops = &btrfs_extent_io_ops;
		inode->i_fop = &btrfs_file_operations;
		inode->i_op = &btrfs_file_inode_operations;
		break;
	case S_IFDIR:
		inode->i_fop = &btrfs_dir_file_operations;
		if (root == root->fs_info->tree_root)
			inode->i_op = &btrfs_dir_ro_inode_operations;
		else
			inode->i_op = &btrfs_dir_inode_operations;
		break;
	case S_IFLNK:
		inode->i_op = &btrfs_symlink_inode_operations;
		inode->i_mapping->a_ops = &btrfs_symlink_aops;
		inode->i_mapping->backing_dev_info = &root->fs_info->bdi;
		break;
	default:
		inode->i_op = &btrfs_special_inode_operations;
		init_special_inode(inode, inode->i_mode, rdev);
		break;
	}

	btrfs_update_iflags(inode);
	return;

make_bad:
	btrfs_free_path(path);
	make_bad_inode(inode);
}

/*
 * given a leaf and an inode, copy the inode fields into the leaf
 */
static void fill_inode_item(struct btrfs_trans_handle *trans,
			    struct extent_buffer *leaf,
			    struct btrfs_inode_item *item,
			    struct inode *inode)
{
	btrfs_set_inode_uid(leaf, item, inode->i_uid);
	btrfs_set_inode_gid(leaf, item, inode->i_gid);
	btrfs_set_inode_size(leaf, item, BTRFS_I(inode)->disk_i_size);
	btrfs_set_inode_mode(leaf, item, inode->i_mode);
	btrfs_set_inode_nlink(leaf, item, inode->i_nlink);

	btrfs_set_timespec_sec(leaf, btrfs_inode_atime(item),
			       inode->i_atime.tv_sec);
	btrfs_set_timespec_nsec(leaf, btrfs_inode_atime(item),
				inode->i_atime.tv_nsec);

	btrfs_set_timespec_sec(leaf, btrfs_inode_mtime(item),
			       inode->i_mtime.tv_sec);
	btrfs_set_timespec_nsec(leaf, btrfs_inode_mtime(item),
				inode->i_mtime.tv_nsec);

	btrfs_set_timespec_sec(leaf, btrfs_inode_ctime(item),
			       inode->i_ctime.tv_sec);
	btrfs_set_timespec_nsec(leaf, btrfs_inode_ctime(item),
				inode->i_ctime.tv_nsec);

	btrfs_set_inode_nbytes(leaf, item, inode_get_bytes(inode));
	btrfs_set_inode_generation(leaf, item, BTRFS_I(inode)->generation);
	btrfs_set_inode_sequence(leaf, item, BTRFS_I(inode)->sequence);
	btrfs_set_inode_transid(leaf, item, trans->transid);
	btrfs_set_inode_rdev(leaf, item, inode->i_rdev);
	btrfs_set_inode_flags(leaf, item, BTRFS_I(inode)->flags);
	btrfs_set_inode_block_group(leaf, item, BTRFS_I(inode)->block_group);
}

/*
 * copy everything in the in-memory inode into the btree.
 */
noinline int btrfs_update_inode(struct btrfs_trans_handle *trans,
				struct btrfs_root *root, struct inode *inode)
{
	struct btrfs_inode_item *inode_item;
	struct btrfs_path *path;
	struct extent_buffer *leaf;
	int ret;

	path = btrfs_alloc_path();
	BUG_ON(!path);
	path->leave_spinning = 1;
	ret = btrfs_lookup_inode(trans, root, path,
				 &BTRFS_I(inode)->location, 1);
	if (ret) {
		if (ret > 0)
			ret = -ENOENT;
		goto failed;
	}

	btrfs_unlock_up_safe(path, 1);
	leaf = path->nodes[0];
	inode_item = btrfs_item_ptr(leaf, path->slots[0],
				  struct btrfs_inode_item);

	fill_inode_item(trans, leaf, inode_item, inode);
	btrfs_mark_buffer_dirty(leaf);
	btrfs_set_inode_last_trans(trans, inode);
	ret = 0;
failed:
	btrfs_free_path(path);
	return ret;
}


/*
 * unlink helper that gets used here in inode.c and in the tree logging
 * recovery code.  It remove a link in a directory with a given name, and
 * also drops the back refs in the inode to the directory
 */
int btrfs_unlink_inode(struct btrfs_trans_handle *trans,
		       struct btrfs_root *root,
		       struct inode *dir, struct inode *inode,
		       const char *name, int name_len)
{
	struct btrfs_path *path;
	int ret = 0;
	struct extent_buffer *leaf;
	struct btrfs_dir_item *di;
	struct btrfs_key key;
	u64 index;

	path = btrfs_alloc_path();
	if (!path) {
		ret = -ENOMEM;
		goto err;
	}

	path->leave_spinning = 1;
	di = btrfs_lookup_dir_item(trans, root, path, dir->i_ino,
				    name, name_len, -1);
	if (IS_ERR(di)) {
		ret = PTR_ERR(di);
		goto err;
	}
	if (!di) {
		ret = -ENOENT;
		goto err;
	}
	leaf = path->nodes[0];
	btrfs_dir_item_key_to_cpu(leaf, di, &key);
	ret = btrfs_delete_one_dir_name(trans, root, path, di);
	if (ret)
		goto err;
	btrfs_release_path(root, path);

	ret = btrfs_del_inode_ref(trans, root, name, name_len,
				  inode->i_ino,
				  dir->i_ino, &index);
	if (ret) {
		printk(KERN_INFO "btrfs failed to delete reference to %.*s, "
		       "inode %lu parent %lu\n", name_len, name,
		       inode->i_ino, dir->i_ino);
		goto err;
	}

	di = btrfs_lookup_dir_index_item(trans, root, path, dir->i_ino,
					 index, name, name_len, -1);
	if (IS_ERR(di)) {
		ret = PTR_ERR(di);
		goto err;
	}
	if (!di) {
		ret = -ENOENT;
		goto err;
	}
	ret = btrfs_delete_one_dir_name(trans, root, path, di);
	btrfs_release_path(root, path);

	ret = btrfs_del_inode_ref_in_log(trans, root, name, name_len,
					 inode, dir->i_ino);
	BUG_ON(ret != 0 && ret != -ENOENT);

	ret = btrfs_del_dir_entries_in_log(trans, root, name, name_len,
					   dir, index);
	BUG_ON(ret);
err:
	btrfs_free_path(path);
	if (ret)
		goto out;

	btrfs_i_size_write(dir, dir->i_size - name_len * 2);
	inode->i_ctime = dir->i_mtime = dir->i_ctime = CURRENT_TIME;
	btrfs_update_inode(trans, root, dir);
	btrfs_drop_nlink(inode);
	ret = btrfs_update_inode(trans, root, inode);
out:
	return ret;
}

static int btrfs_unlink(struct inode *dir, struct dentry *dentry)
{
	struct btrfs_root *root;
	struct btrfs_trans_handle *trans;
	struct inode *inode = dentry->d_inode;
	int ret;
	unsigned long nr = 0;

	root = BTRFS_I(dir)->root;

	trans = btrfs_start_transaction(root, 1);

	btrfs_set_trans_block_group(trans, dir);

	btrfs_record_unlink_dir(trans, dir, dentry->d_inode, 0);

	ret = btrfs_unlink_inode(trans, root, dir, dentry->d_inode,
				 dentry->d_name.name, dentry->d_name.len);

	if (inode->i_nlink == 0)
		ret = btrfs_orphan_add(trans, inode);

	nr = trans->blocks_used;

	btrfs_end_transaction_throttle(trans, root);
	btrfs_btree_balance_dirty(root, nr);
	return ret;
}

static int btrfs_rmdir(struct inode *dir, struct dentry *dentry)
{
	struct inode *inode = dentry->d_inode;
	int err = 0;
	int ret;
	struct btrfs_root *root = BTRFS_I(dir)->root;
	struct btrfs_trans_handle *trans;
	unsigned long nr = 0;

	/*
	 * the FIRST_FREE_OBJECTID check makes sure we don't try to rmdir
	 * the root of a subvolume or snapshot
	 */
	if (inode->i_size > BTRFS_EMPTY_DIR_SIZE ||
	    inode->i_ino == BTRFS_FIRST_FREE_OBJECTID) {
		return -ENOTEMPTY;
	}

	trans = btrfs_start_transaction(root, 1);
	btrfs_set_trans_block_group(trans, dir);

	err = btrfs_orphan_add(trans, inode);
	if (err)
		goto fail_trans;

	/* now the directory is empty */
	err = btrfs_unlink_inode(trans, root, dir, dentry->d_inode,
				 dentry->d_name.name, dentry->d_name.len);
	if (!err)
		btrfs_i_size_write(inode, 0);

fail_trans:
	nr = trans->blocks_used;
	ret = btrfs_end_transaction_throttle(trans, root);
	btrfs_btree_balance_dirty(root, nr);

	if (ret && !err)
		err = ret;
	return err;
}

#if 0
/*
 * when truncating bytes in a file, it is possible to avoid reading
 * the leaves that contain only checksum items.  This can be the
 * majority of the IO required to delete a large file, but it must
 * be done carefully.
 *
 * The keys in the level just above the leaves are checked to make sure
 * the lowest key in a given leaf is a csum key, and starts at an offset
 * after the new  size.
 *
 * Then the key for the next leaf is checked to make sure it also has
 * a checksum item for the same file.  If it does, we know our target leaf
 * contains only checksum items, and it can be safely freed without reading
 * it.
 *
 * This is just an optimization targeted at large files.  It may do
 * nothing.  It will return 0 unless things went badly.
 */
static noinline int drop_csum_leaves(struct btrfs_trans_handle *trans,
				     struct btrfs_root *root,
				     struct btrfs_path *path,
				     struct inode *inode, u64 new_size)
{
	struct btrfs_key key;
	int ret;
	int nritems;
	struct btrfs_key found_key;
	struct btrfs_key other_key;
	struct btrfs_leaf_ref *ref;
	u64 leaf_gen;
	u64 leaf_start;

	path->lowest_level = 1;
	key.objectid = inode->i_ino;
	key.type = BTRFS_CSUM_ITEM_KEY;
	key.offset = new_size;
again:
	ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
	if (ret < 0)
		goto out;

	if (path->nodes[1] == NULL) {
		ret = 0;
		goto out;
	}
	ret = 0;
	btrfs_node_key_to_cpu(path->nodes[1], &found_key, path->slots[1]);
	nritems = btrfs_header_nritems(path->nodes[1]);

	if (!nritems)
		goto out;

	if (path->slots[1] >= nritems)
		goto next_node;

	/* did we find a key greater than anything we want to delete? */
	if (found_key.objectid > inode->i_ino ||
	   (found_key.objectid == inode->i_ino && found_key.type > key.type))
		goto out;

	/* we check the next key in the node to make sure the leave contains
	 * only checksum items.  This comparison doesn't work if our
	 * leaf is the last one in the node
	 */
	if (path->slots[1] + 1 >= nritems) {
next_node:
		/* search forward from the last key in the node, this
		 * will bring us into the next node in the tree
		 */
		btrfs_node_key_to_cpu(path->nodes[1], &found_key, nritems - 1);

		/* unlikely, but we inc below, so check to be safe */
		if (found_key.offset == (u64)-1)
			goto out;

		/* search_forward needs a path with locks held, do the
		 * search again for the original key.  It is possible
		 * this will race with a balance and return a path that
		 * we could modify, but this drop is just an optimization
		 * and is allowed to miss some leaves.
		 */
		btrfs_release_path(root, path);
		found_key.offset++;

		/* setup a max key for search_forward */
		other_key.offset = (u64)-1;
		other_key.type = key.type;
		other_key.objectid = key.objectid;

		path->keep_locks = 1;
		ret = btrfs_search_forward(root, &found_key, &other_key,
					   path, 0, 0);
		path->keep_locks = 0;
		if (ret || found_key.objectid != key.objectid ||
		    found_key.type != key.type) {
			ret = 0;
			goto out;
		}

		key.offset = found_key.offset;
		btrfs_release_path(root, path);
		cond_resched();
		goto again;
	}

	/* we know there's one more slot after us in the tree,
	 * read that key so we can verify it is also a checksum item
	 */
	btrfs_node_key_to_cpu(path->nodes[1], &other_key, path->slots[1] + 1);

	if (found_key.objectid < inode->i_ino)
		goto next_key;

	if (found_key.type != key.type || found_key.offset < new_size)
		goto next_key;

	/*
	 * if the key for the next leaf isn't a csum key from this objectid,
	 * we can't be sure there aren't good items inside this leaf.
	 * Bail out
	 */
	if (other_key.objectid != inode->i_ino || other_key.type != key.type)
		goto out;

	leaf_start = btrfs_node_blockptr(path->nodes[1], path->slots[1]);
	leaf_gen = btrfs_node_ptr_generation(path->nodes[1], path->slots[1]);
	/*
	 * it is safe to delete this leaf, it contains only
	 * csum items from this inode at an offset >= new_size
	 */
	ret = btrfs_del_leaf(trans, root, path, leaf_start);
	BUG_ON(ret);

	if (root->ref_cows && leaf_gen < trans->transid) {
		ref = btrfs_alloc_leaf_ref(root, 0);
		if (ref) {
			ref->root_gen = root->root_key.offset;
			ref->bytenr = leaf_start;
			ref->owner = 0;
			ref->generation = leaf_gen;
			ref->nritems = 0;

			btrfs_sort_leaf_ref(ref);

			ret = btrfs_add_leaf_ref(root, ref, 0);
			WARN_ON(ret);
			btrfs_free_leaf_ref(root, ref);
		} else {
			WARN_ON(1);
		}
	}
next_key:
	btrfs_release_path(root, path);

	if (other_key.objectid == inode->i_ino &&
	    other_key.type == key.type && other_key.offset > key.offset) {
		key.offset = other_key.offset;
		cond_resched();
		goto again;
	}
	ret = 0;
out:
	/* fixup any changes we've made to the path */
	path->lowest_level = 0;
	path->keep_locks = 0;
	btrfs_release_path(root, path);
	return ret;
}

#endif

/*
 * this can truncate away extent items, csum items and directory items.
 * It starts at a high offset and removes keys until it can't find
 * any higher than new_size
 *
 * csum items that cross the new i_size are truncated to the new size
 * as well.
 *
 * min_type is the minimum key type to truncate down to.  If set to 0, this
 * will kill all the items on this inode, including the INODE_ITEM_KEY.
 */
noinline int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans,
					struct btrfs_root *root,
					struct inode *inode,
					u64 new_size, u32 min_type)
{
	int ret;
	struct btrfs_path *path;
	struct btrfs_key key;
	struct btrfs_key found_key;
	u32 found_type = (u8)-1;
	struct extent_buffer *leaf;
	struct btrfs_file_extent_item *fi;
	u64 extent_start = 0;
	u64 extent_num_bytes = 0;
	u64 extent_offset = 0;
	u64 item_end = 0;
	int found_extent;
	int del_item;
	int pending_del_nr = 0;
	int pending_del_slot = 0;
	int extent_type = -1;
	int encoding;
	u64 mask = root->sectorsize - 1;

	if (root->ref_cows)
		btrfs_drop_extent_cache(inode, new_size & (~mask), (u64)-1, 0);
	path = btrfs_alloc_path();
	BUG_ON(!path);
	path->reada = -1;

	/* FIXME, add redo link to tree so we don't leak on crash */
	key.objectid = inode->i_ino;
	key.offset = (u64)-1;
	key.type = (u8)-1;

search_again:
	path->leave_spinning = 1;
	ret = btrfs_search_slot(trans, root, &key, path, -1, 1);
	if (ret < 0)
		goto error;

	if (ret > 0) {
		/* there are no items in the tree for us to truncate, we're
		 * done
		 */
		if (path->slots[0] == 0) {
			ret = 0;
			goto error;
		}
		path->slots[0]--;
	}

	while (1) {
		fi = NULL;
		leaf = path->nodes[0];
		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
		found_type = btrfs_key_type(&found_key);
		encoding = 0;

		if (found_key.objectid != inode->i_ino)
			break;

		if (found_type < min_type)
			break;

		item_end = found_key.offset;
		if (found_type == BTRFS_EXTENT_DATA_KEY) {
			fi = btrfs_item_ptr(leaf, path->slots[0],
					    struct btrfs_file_extent_item);
			extent_type = btrfs_file_extent_type(leaf, fi);
			encoding = btrfs_file_extent_compression(leaf, fi);
			encoding |= btrfs_file_extent_encryption(leaf, fi);
			encoding |= btrfs_file_extent_other_encoding(leaf, fi);

			if (extent_type != BTRFS_FILE_EXTENT_INLINE) {
				item_end +=
				    btrfs_file_extent_num_bytes(leaf, fi);
			} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
				item_end += btrfs_file_extent_inline_len(leaf,
									 fi);
			}
			item_end--;
		}
		if (item_end < new_size) {
			if (found_type == BTRFS_DIR_ITEM_KEY)
				found_type = BTRFS_INODE_ITEM_KEY;
			else if (found_type == BTRFS_EXTENT_ITEM_KEY)
				found_type = BTRFS_EXTENT_DATA_KEY;
			else if (found_type == BTRFS_EXTENT_DATA_KEY)
				found_type = BTRFS_XATTR_ITEM_KEY;
			else if (found_type == BTRFS_XATTR_ITEM_KEY)
				found_type = BTRFS_INODE_REF_KEY;
			else if (found_type)
				found_type--;
			else
				break;
			btrfs_set_key_type(&key, found_type);
			goto next;
		}
		if (found_key.offset >= new_size)
			del_item = 1;
		else
			del_item = 0;
		found_extent = 0;

		/* FIXME, shrink the extent if the ref count is only 1 */
		if (found_type != BTRFS_EXTENT_DATA_KEY)
			goto delete;

		if (extent_type != BTRFS_FILE_EXTENT_INLINE) {
			u64 num_dec;
			extent_start = btrfs_file_extent_disk_bytenr(leaf, fi);
			if (!del_item && !encoding) {
				u64 orig_num_bytes =
					btrfs_file_extent_num_bytes(leaf, fi);
				extent_num_bytes = new_size -
					found_key.offset + root->sectorsize - 1;
				extent_num_bytes = extent_num_bytes &
					~((u64)root->sectorsize - 1);
				btrfs_set_file_extent_num_bytes(leaf, fi,
							 extent_num_bytes);
				num_dec = (orig_num_bytes -
					   extent_num_bytes);
				if (root->ref_cows && extent_start != 0)
					inode_sub_bytes(inode, num_dec);
				btrfs_mark_buffer_dirty(leaf);
			} else {
				extent_num_bytes =
					btrfs_file_extent_disk_num_bytes(leaf,
									 fi);
				extent_offset = found_key.offset -
					btrfs_file_extent_offset(leaf, fi);

				/* FIXME blocksize != 4096 */
				num_dec = btrfs_file_extent_num_bytes(leaf, fi);
				if (extent_start != 0) {
					found_extent = 1;
					if (root->ref_cows)
						inode_sub_bytes(inode, num_dec);
				}
			}
		} else if (extent_type == BTRFS_FILE_EXTENT_INLINE) {
			/*
			 * we can't truncate inline items that have had
			 * special encodings
			 */
			if (!del_item &&
			    btrfs_file_extent_compression(leaf, fi) == 0 &&
			    btrfs_file_extent_encryption(leaf, fi) == 0 &&
			    btrfs_file_extent_other_encoding(leaf, fi) == 0) {
				u32 size = new_size - found_key.offset;

				if (root->ref_cows) {
					inode_sub_bytes(inode, item_end + 1 -
							new_size);
				}
				size =
				    btrfs_file_extent_calc_inline_size(size);
				ret = btrfs_truncate_item(trans, root, path,
							  size, 1);
				BUG_ON(ret);
			} else if (root->ref_cows) {
				inode_sub_bytes(inode, item_end + 1 -
						found_key.offset);
			}
		}
delete:
		if (del_item) {
			if (!pending_del_nr) {
				/* no pending yet, add ourselves */
				pending_del_slot = path->slots[0];
				pending_del_nr = 1;
			} else if (pending_del_nr &&
				   path->slots[0] + 1 == pending_del_slot) {
				/* hop on the pending chunk */
				pending_del_nr++;
				pending_del_slot = path->slots[0];
			} else {
				BUG();
			}
		} else {
			break;
		}
		if (found_extent && root->ref_cows) {
			btrfs_set_path_blocking(path);
			ret = btrfs_free_extent(trans, root, extent_start,
						extent_num_bytes, 0,
						btrfs_header_owner(leaf),
						inode->i_ino, extent_offset);
			BUG_ON(ret);
		}
next:
		if (path->slots[0] == 0) {
			if (pending_del_nr)
				goto del_pending;
			btrfs_release_path(root, path);
			if (found_type == BTRFS_INODE_ITEM_KEY)
				break;
			goto search_again;
		}

		path->slots[0]--;
		if (pending_del_nr &&
		    path->slots[0] + 1 != pending_del_slot) {
			struct btrfs_key debug;
del_pending:
			btrfs_item_key_to_cpu(path->nodes[0], &debug,
					      pending_del_slot);
			ret = btrfs_del_items(trans, root, path,
					      pending_del_slot,
					      pending_del_nr);
			BUG_ON(ret);
			pending_del_nr = 0;
			btrfs_release_path(root, path);
			if (found_type == BTRFS_INODE_ITEM_KEY)
				break;
			goto search_again;
		}
	}
	ret = 0;
error:
	if (pending_del_nr) {
		ret = btrfs_del_items(trans, root, path, pending_del_slot,
				      pending_del_nr);
	}
	btrfs_free_path(path);
	return ret;
}

/*
 * taken from block_truncate_page, but does cow as it zeros out
 * any bytes left in the last page in the file.
 */
static int btrfs_truncate_page(struct address_space *mapping, loff_t from)
{
	struct inode *inode = mapping->host;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct btrfs_ordered_extent *ordered;
	char *kaddr;
	u32 blocksize = root->sectorsize;
	pgoff_t index = from >> PAGE_CACHE_SHIFT;
	unsigned offset = from & (PAGE_CACHE_SIZE-1);
	struct page *page;
	int ret = 0;
	u64 page_start;
	u64 page_end;

	if ((offset & (blocksize - 1)) == 0)
		goto out;

	ret = -ENOMEM;
again:
	page = grab_cache_page(mapping, index);
	if (!page)
		goto out;

	page_start = page_offset(page);
	page_end = page_start + PAGE_CACHE_SIZE - 1;

	if (!PageUptodate(page)) {
		ret = btrfs_readpage(NULL, page);
		lock_page(page);
		if (page->mapping != mapping) {
			unlock_page(page);
			page_cache_release(page);
			goto again;
		}
		if (!PageUptodate(page)) {
			ret = -EIO;
			goto out_unlock;
		}
	}
	wait_on_page_writeback(page);

	lock_extent(io_tree, page_start, page_end, GFP_NOFS);
	set_page_extent_mapped(page);

	ordered = btrfs_lookup_ordered_extent(inode, page_start);
	if (ordered) {
		unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
		unlock_page(page);
		page_cache_release(page);
		btrfs_start_ordered_extent(inode, ordered, 1);
		btrfs_put_ordered_extent(ordered);
		goto again;
	}

	btrfs_set_extent_delalloc(inode, page_start, page_end);
	ret = 0;
	if (offset != PAGE_CACHE_SIZE) {
		kaddr = kmap(page);
		memset(kaddr + offset, 0, PAGE_CACHE_SIZE - offset);
		flush_dcache_page(page);
		kunmap(page);
	}
	ClearPageChecked(page);
	set_page_dirty(page);
	unlock_extent(io_tree, page_start, page_end, GFP_NOFS);

out_unlock:
	unlock_page(page);
	page_cache_release(page);
out:
	return ret;
}

int btrfs_cont_expand(struct inode *inode, loff_t size)
{
	struct btrfs_trans_handle *trans;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct extent_map *em;
	u64 mask = root->sectorsize - 1;
	u64 hole_start = (inode->i_size + mask) & ~mask;
	u64 block_end = (size + mask) & ~mask;
	u64 last_byte;
	u64 cur_offset;
	u64 hole_size;
	int err;

	if (size <= hole_start)
		return 0;

	err = btrfs_check_metadata_free_space(root);
	if (err)
		return err;

	btrfs_truncate_page(inode->i_mapping, inode->i_size);

	while (1) {
		struct btrfs_ordered_extent *ordered;
		btrfs_wait_ordered_range(inode, hole_start,
					 block_end - hole_start);
		lock_extent(io_tree, hole_start, block_end - 1, GFP_NOFS);
		ordered = btrfs_lookup_ordered_extent(inode, hole_start);
		if (!ordered)
			break;
		unlock_extent(io_tree, hole_start, block_end - 1, GFP_NOFS);
		btrfs_put_ordered_extent(ordered);
	}

	trans = btrfs_start_transaction(root, 1);
	btrfs_set_trans_block_group(trans, inode);

	cur_offset = hole_start;
	while (1) {
		em = btrfs_get_extent(inode, NULL, 0, cur_offset,
				block_end - cur_offset, 0);
		BUG_ON(IS_ERR(em) || !em);
		last_byte = min(extent_map_end(em), block_end);
		last_byte = (last_byte + mask) & ~mask;
		if (test_bit(EXTENT_FLAG_VACANCY, &em->flags)) {
			u64 hint_byte = 0;
			hole_size = last_byte - cur_offset;
			err = btrfs_drop_extents(trans, root, inode,
						 cur_offset,
						 cur_offset + hole_size,
						 block_end,
						 cur_offset, &hint_byte);
			if (err)
				break;
			err = btrfs_insert_file_extent(trans, root,
					inode->i_ino, cur_offset, 0,
					0, hole_size, 0, hole_size,
					0, 0, 0);
			btrfs_drop_extent_cache(inode, hole_start,
					last_byte - 1, 0);
		}
		free_extent_map(em);
		cur_offset = last_byte;
		if (err || cur_offset >= block_end)
			break;
	}

	btrfs_end_transaction(trans, root);
	unlock_extent(io_tree, hole_start, block_end - 1, GFP_NOFS);
	return err;
}

static int btrfs_setattr(struct dentry *dentry, struct iattr *attr)
{
	struct inode *inode = dentry->d_inode;
	int err;

	err = inode_change_ok(inode, attr);
	if (err)
		return err;

	if (S_ISREG(inode->i_mode) && (attr->ia_valid & ATTR_SIZE)) {
		if (attr->ia_size > inode->i_size) {
			err = btrfs_cont_expand(inode, attr->ia_size);
			if (err)
				return err;
		} else if (inode->i_size > 0 &&
			   attr->ia_size == 0) {

			/* we're truncating a file that used to have good
			 * data down to zero.  Make sure it gets into
			 * the ordered flush list so that any new writes
			 * get down to disk quickly.
			 */
			BTRFS_I(inode)->ordered_data_close = 1;
		}
	}

	err = inode_setattr(inode, attr);

	if (!err && ((attr->ia_valid & ATTR_MODE)))
		err = btrfs_acl_chmod(inode);
	return err;
}

void btrfs_delete_inode(struct inode *inode)
{
	struct btrfs_trans_handle *trans;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	unsigned long nr;
	int ret;

	truncate_inode_pages(&inode->i_data, 0);
	if (is_bad_inode(inode)) {
		btrfs_orphan_del(NULL, inode);
		goto no_delete;
	}
	btrfs_wait_ordered_range(inode, 0, (u64)-1);

	btrfs_i_size_write(inode, 0);
	trans = btrfs_join_transaction(root, 1);

	btrfs_set_trans_block_group(trans, inode);
	ret = btrfs_truncate_inode_items(trans, root, inode, inode->i_size, 0);
	if (ret) {
		btrfs_orphan_del(NULL, inode);
		goto no_delete_lock;
	}

	btrfs_orphan_del(trans, inode);

	nr = trans->blocks_used;
	clear_inode(inode);

	btrfs_end_transaction(trans, root);
	btrfs_btree_balance_dirty(root, nr);
	return;

no_delete_lock:
	nr = trans->blocks_used;
	btrfs_end_transaction(trans, root);
	btrfs_btree_balance_dirty(root, nr);
no_delete:
	clear_inode(inode);
}

/*
 * this returns the key found in the dir entry in the location pointer.
 * If no dir entries were found, location->objectid is 0.
 */
static int btrfs_inode_by_name(struct inode *dir, struct dentry *dentry,
			       struct btrfs_key *location)
{
	const char *name = dentry->d_name.name;
	int namelen = dentry->d_name.len;
	struct btrfs_dir_item *di;
	struct btrfs_path *path;
	struct btrfs_root *root = BTRFS_I(dir)->root;
	int ret = 0;

	path = btrfs_alloc_path();
	BUG_ON(!path);

	di = btrfs_lookup_dir_item(NULL, root, path, dir->i_ino, name,
				    namelen, 0);
	if (IS_ERR(di))
		ret = PTR_ERR(di);

	if (!di || IS_ERR(di))
		goto out_err;

	btrfs_dir_item_key_to_cpu(path->nodes[0], di, location);
out:
	btrfs_free_path(path);
	return ret;
out_err:
	location->objectid = 0;
	goto out;
}

/*
 * when we hit a tree root in a directory, the btrfs part of the inode
 * needs to be changed to reflect the root directory of the tree root.  This
 * is kind of like crossing a mount point.
 */
static int fixup_tree_root_location(struct btrfs_root *root,
			     struct btrfs_key *location,
			     struct btrfs_root **sub_root,
			     struct dentry *dentry)
{
	struct btrfs_root_item *ri;

	if (btrfs_key_type(location) != BTRFS_ROOT_ITEM_KEY)
		return 0;
	if (location->objectid == BTRFS_ROOT_TREE_OBJECTID)
		return 0;

	*sub_root = btrfs_read_fs_root(root->fs_info, location,
					dentry->d_name.name,
					dentry->d_name.len);
	if (IS_ERR(*sub_root))
		return PTR_ERR(*sub_root);

	ri = &(*sub_root)->root_item;
	location->objectid = btrfs_root_dirid(ri);
	btrfs_set_key_type(location, BTRFS_INODE_ITEM_KEY);
	location->offset = 0;

	return 0;
}

static void inode_tree_add(struct inode *inode)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_inode *entry;
	struct rb_node **p;
	struct rb_node *parent;

again:
	p = &root->inode_tree.rb_node;
	parent = NULL;

	spin_lock(&root->inode_lock);
	while (*p) {
		parent = *p;
		entry = rb_entry(parent, struct btrfs_inode, rb_node);

		if (inode->i_ino < entry->vfs_inode.i_ino)
			p = &parent->rb_left;
		else if (inode->i_ino > entry->vfs_inode.i_ino)
			p = &parent->rb_right;
		else {
			WARN_ON(!(entry->vfs_inode.i_state &
				  (I_WILL_FREE | I_FREEING | I_CLEAR)));
			rb_erase(parent, &root->inode_tree);
			RB_CLEAR_NODE(parent);
			spin_unlock(&root->inode_lock);
			goto again;
		}
	}
	rb_link_node(&BTRFS_I(inode)->rb_node, parent, p);
	rb_insert_color(&BTRFS_I(inode)->rb_node, &root->inode_tree);
	spin_unlock(&root->inode_lock);
}

static void inode_tree_del(struct inode *inode)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;

	spin_lock(&root->inode_lock);
	if (!RB_EMPTY_NODE(&BTRFS_I(inode)->rb_node)) {
		rb_erase(&BTRFS_I(inode)->rb_node, &root->inode_tree);
		RB_CLEAR_NODE(&BTRFS_I(inode)->rb_node);
	}
	spin_unlock(&root->inode_lock);
}

static noinline void init_btrfs_i(struct inode *inode)
{
	struct btrfs_inode *bi = BTRFS_I(inode);

	bi->generation = 0;
	bi->sequence = 0;
	bi->last_trans = 0;
	bi->logged_trans = 0;
	bi->delalloc_bytes = 0;
	bi->reserved_bytes = 0;
	bi->disk_i_size = 0;
	bi->flags = 0;
	bi->index_cnt = (u64)-1;
	bi->last_unlink_trans = 0;
	bi->ordered_data_close = 0;
	extent_map_tree_init(&BTRFS_I(inode)->extent_tree, GFP_NOFS);
	extent_io_tree_init(&BTRFS_I(inode)->io_tree,
			     inode->i_mapping, GFP_NOFS);
	extent_io_tree_init(&BTRFS_I(inode)->io_failure_tree,
			     inode->i_mapping, GFP_NOFS);
	INIT_LIST_HEAD(&BTRFS_I(inode)->delalloc_inodes);
	INIT_LIST_HEAD(&BTRFS_I(inode)->ordered_operations);
	RB_CLEAR_NODE(&BTRFS_I(inode)->rb_node);
	btrfs_ordered_inode_tree_init(&BTRFS_I(inode)->ordered_tree);
	mutex_init(&BTRFS_I(inode)->extent_mutex);
	mutex_init(&BTRFS_I(inode)->log_mutex);
}

static int btrfs_init_locked_inode(struct inode *inode, void *p)
{
	struct btrfs_iget_args *args = p;
	inode->i_ino = args->ino;
	init_btrfs_i(inode);
	BTRFS_I(inode)->root = args->root;
	btrfs_set_inode_space_info(args->root, inode);
	return 0;
}

static int btrfs_find_actor(struct inode *inode, void *opaque)
{
	struct btrfs_iget_args *args = opaque;
	return args->ino == inode->i_ino &&
		args->root == BTRFS_I(inode)->root;
}

static struct inode *btrfs_iget_locked(struct super_block *s,
				       u64 objectid,
				       struct btrfs_root *root)
{
	struct inode *inode;
	struct btrfs_iget_args args;
	args.ino = objectid;
	args.root = root;

	inode = iget5_locked(s, objectid, btrfs_find_actor,
			     btrfs_init_locked_inode,
			     (void *)&args);
	return inode;
}

/* Get an inode object given its location and corresponding root.
 * Returns in *is_new if the inode was read from disk
 */
struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location,
			 struct btrfs_root *root)
{
	struct inode *inode;

	inode = btrfs_iget_locked(s, location->objectid, root);
	if (!inode)
		return ERR_PTR(-ENOMEM);

	if (inode->i_state & I_NEW) {
		BTRFS_I(inode)->root = root;
		memcpy(&BTRFS_I(inode)->location, location, sizeof(*location));
		btrfs_read_locked_inode(inode);

		inode_tree_add(inode);
		unlock_new_inode(inode);
	}

	return inode;
}

struct inode *btrfs_lookup_dentry(struct inode *dir, struct dentry *dentry)
{
	struct inode *inode;
	struct btrfs_inode *bi = BTRFS_I(dir);
	struct btrfs_root *root = bi->root;
	struct btrfs_root *sub_root = root;
	struct btrfs_key location;
	int ret;

	if (dentry->d_name.len > BTRFS_NAME_LEN)
		return ERR_PTR(-ENAMETOOLONG);

	ret = btrfs_inode_by_name(dir, dentry, &location);

	if (ret < 0)
		return ERR_PTR(ret);

	inode = NULL;
	if (location.objectid) {
		ret = fixup_tree_root_location(root, &location, &sub_root,
						dentry);
		if (ret < 0)
			return ERR_PTR(ret);
		if (ret > 0)
			return ERR_PTR(-ENOENT);
		inode = btrfs_iget(dir->i_sb, &location, sub_root);
		if (IS_ERR(inode))
			return ERR_CAST(inode);
	}
	return inode;
}

static struct dentry *btrfs_lookup(struct inode *dir, struct dentry *dentry,
				   struct nameidata *nd)
{
	struct inode *inode;

	if (dentry->d_name.len > BTRFS_NAME_LEN)
		return ERR_PTR(-ENAMETOOLONG);

	inode = btrfs_lookup_dentry(dir, dentry);
	if (IS_ERR(inode))
		return ERR_CAST(inode);

	return d_splice_alias(inode, dentry);
}

static unsigned char btrfs_filetype_table[] = {
	DT_UNKNOWN, DT_REG, DT_DIR, DT_CHR, DT_BLK, DT_FIFO, DT_SOCK, DT_LNK
};

static int btrfs_real_readdir(struct file *filp, void *dirent,
			      filldir_t filldir)
{
	struct inode *inode = filp->f_dentry->d_inode;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_item *item;
	struct btrfs_dir_item *di;
	struct btrfs_key key;
	struct btrfs_key found_key;
	struct btrfs_path *path;
	int ret;
	u32 nritems;
	struct extent_buffer *leaf;
	int slot;
	int advance;
	unsigned char d_type;
	int over = 0;
	u32 di_cur;
	u32 di_total;
	u32 di_len;
	int key_type = BTRFS_DIR_INDEX_KEY;
	char tmp_name[32];
	char *name_ptr;
	int name_len;

	/* FIXME, use a real flag for deciding about the key type */
	if (root->fs_info->tree_root == root)
		key_type = BTRFS_DIR_ITEM_KEY;

	/* special case for "." */
	if (filp->f_pos == 0) {
		over = filldir(dirent, ".", 1,
			       1, inode->i_ino,
			       DT_DIR);
		if (over)
			return 0;
		filp->f_pos = 1;
	}
	/* special case for .., just use the back ref */
	if (filp->f_pos == 1) {
		u64 pino = parent_ino(filp->f_path.dentry);
		over = filldir(dirent, "..", 2,
			       2, pino, DT_DIR);
		if (over)
			return 0;
		filp->f_pos = 2;
	}
	path = btrfs_alloc_path();
	path->reada = 2;

	btrfs_set_key_type(&key, key_type);
	key.offset = filp->f_pos;
	key.objectid = inode->i_ino;

	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret < 0)
		goto err;
	advance = 0;

	while (1) {
		leaf = path->nodes[0];
		nritems = btrfs_header_nritems(leaf);
		slot = path->slots[0];
		if (advance || slot >= nritems) {
			if (slot >= nritems - 1) {
				ret = btrfs_next_leaf(root, path);
				if (ret)
					break;
				leaf = path->nodes[0];
				nritems = btrfs_header_nritems(leaf);
				slot = path->slots[0];
			} else {
				slot++;
				path->slots[0]++;
			}
		}

		advance = 1;
		item = btrfs_item_nr(leaf, slot);
		btrfs_item_key_to_cpu(leaf, &found_key, slot);

		if (found_key.objectid != key.objectid)
			break;
		if (btrfs_key_type(&found_key) != key_type)
			break;
		if (found_key.offset < filp->f_pos)
			continue;

		filp->f_pos = found_key.offset;

		di = btrfs_item_ptr(leaf, slot, struct btrfs_dir_item);
		di_cur = 0;
		di_total = btrfs_item_size(leaf, item);

		while (di_cur < di_total) {
			struct btrfs_key location;

			name_len = btrfs_dir_name_len(leaf, di);
			if (name_len <= sizeof(tmp_name)) {
				name_ptr = tmp_name;
			} else {
				name_ptr = kmalloc(name_len, GFP_NOFS);
				if (!name_ptr) {
					ret = -ENOMEM;
					goto err;
				}
			}
			read_extent_buffer(leaf, name_ptr,
					   (unsigned long)(di + 1), name_len);

			d_type = btrfs_filetype_table[btrfs_dir_type(leaf, di)];
			btrfs_dir_item_key_to_cpu(leaf, di, &location);

			/* is this a reference to our own snapshot? If so
			 * skip it
			 */
			if (location.type == BTRFS_ROOT_ITEM_KEY &&
			    location.objectid == root->root_key.objectid) {
				over = 0;
				goto skip;
			}
			over = filldir(dirent, name_ptr, name_len,
				       found_key.offset, location.objectid,
				       d_type);

skip:
			if (name_ptr != tmp_name)
				kfree(name_ptr);

			if (over)
				goto nopos;
			di_len = btrfs_dir_name_len(leaf, di) +
				 btrfs_dir_data_len(leaf, di) + sizeof(*di);
			di_cur += di_len;
			di = (struct btrfs_dir_item *)((char *)di + di_len);
		}
	}

	/* Reached end of directory/root. Bump pos past the last item. */
	if (key_type == BTRFS_DIR_INDEX_KEY)
		filp->f_pos = INT_LIMIT(off_t);
	else
		filp->f_pos++;
nopos:
	ret = 0;
err:
	btrfs_free_path(path);
	return ret;
}

int btrfs_write_inode(struct inode *inode, int wait)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;
	int ret = 0;

	if (root->fs_info->btree_inode == inode)
		return 0;

	if (wait) {
		trans = btrfs_join_transaction(root, 1);
		btrfs_set_trans_block_group(trans, inode);
		ret = btrfs_commit_transaction(trans, root);
	}
	return ret;
}

/*
 * This is somewhat expensive, updating the tree every time the
 * inode changes.  But, it is most likely to find the inode in cache.
 * FIXME, needs more benchmarking...there are no reasons other than performance
 * to keep or drop this code.
 */
void btrfs_dirty_inode(struct inode *inode)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_trans_handle *trans;

	trans = btrfs_join_transaction(root, 1);
	btrfs_set_trans_block_group(trans, inode);
	btrfs_update_inode(trans, root, inode);
	btrfs_end_transaction(trans, root);
}

/*
 * find the highest existing sequence number in a directory
 * and then set the in-memory index_cnt variable to reflect
 * free sequence numbers
 */
static int btrfs_set_inode_index_count(struct inode *inode)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_key key, found_key;
	struct btrfs_path *path;
	struct extent_buffer *leaf;
	int ret;

	key.objectid = inode->i_ino;
	btrfs_set_key_type(&key, BTRFS_DIR_INDEX_KEY);
	key.offset = (u64)-1;

	path = btrfs_alloc_path();
	if (!path)
		return -ENOMEM;

	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
	if (ret < 0)
		goto out;
	/* FIXME: we should be able to handle this */
	if (ret == 0)
		goto out;
	ret = 0;

	/*
	 * MAGIC NUMBER EXPLANATION:
	 * since we search a directory based on f_pos we have to start at 2
	 * since '.' and '..' have f_pos of 0 and 1 respectively, so everybody
	 * else has to start at 2
	 */
	if (path->slots[0] == 0) {
		BTRFS_I(inode)->index_cnt = 2;
		goto out;
	}

	path->slots[0]--;

	leaf = path->nodes[0];
	btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);

	if (found_key.objectid != inode->i_ino ||
	    btrfs_key_type(&found_key) != BTRFS_DIR_INDEX_KEY) {
		BTRFS_I(inode)->index_cnt = 2;
		goto out;
	}

	BTRFS_I(inode)->index_cnt = found_key.offset + 1;
out:
	btrfs_free_path(path);
	return ret;
}

/*
 * helper to find a free sequence number in a given directory.  This current
 * code is very simple, later versions will do smarter things in the btree
 */
int btrfs_set_inode_index(struct inode *dir, u64 *index)
{
	int ret = 0;

	if (BTRFS_I(dir)->index_cnt == (u64)-1) {
		ret = btrfs_set_inode_index_count(dir);
		if (ret)
			return ret;
	}

	*index = BTRFS_I(dir)->index_cnt;
	BTRFS_I(dir)->index_cnt++;

	return ret;
}

static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans,
				     struct btrfs_root *root,
				     struct inode *dir,
				     const char *name, int name_len,
				     u64 ref_objectid, u64 objectid,
				     u64 alloc_hint, int mode, u64 *index)
{
	struct inode *inode;
	struct btrfs_inode_item *inode_item;
	struct btrfs_key *location;
	struct btrfs_path *path;
	struct btrfs_inode_ref *ref;
	struct btrfs_key key[2];
	u32 sizes[2];
	unsigned long ptr;
	int ret;
	int owner;

	path = btrfs_alloc_path();
	BUG_ON(!path);

	inode = new_inode(root->fs_info->sb);
	if (!inode)
		return ERR_PTR(-ENOMEM);

	if (dir) {
		ret = btrfs_set_inode_index(dir, index);
		if (ret) {
			iput(inode);
			return ERR_PTR(ret);
		}
	}
	/*
	 * index_cnt is ignored for everything but a dir,
	 * btrfs_get_inode_index_count has an explanation for the magic
	 * number
	 */
	init_btrfs_i(inode);
	BTRFS_I(inode)->index_cnt = 2;
	BTRFS_I(inode)->root = root;
	BTRFS_I(inode)->generation = trans->transid;
	btrfs_set_inode_space_info(root, inode);

	if (mode & S_IFDIR)
		owner = 0;
	else
		owner = 1;
	BTRFS_I(inode)->block_group =
			btrfs_find_block_group(root, 0, alloc_hint, owner);

	key[0].objectid = objectid;
	btrfs_set_key_type(&key[0], BTRFS_INODE_ITEM_KEY);
	key[0].offset = 0;

	key[1].objectid = objectid;
	btrfs_set_key_type(&key[1], BTRFS_INODE_REF_KEY);
	key[1].offset = ref_objectid;

	sizes[0] = sizeof(struct btrfs_inode_item);
	sizes[1] = name_len + sizeof(*ref);

	path->leave_spinning = 1;
	ret = btrfs_insert_empty_items(trans, root, path, key, sizes, 2);
	if (ret != 0)
		goto fail;

	if (objectid > root->highest_inode)
		root->highest_inode = objectid;

	inode->i_uid = current_fsuid();

	if (dir && (dir->i_mode & S_ISGID)) {
		inode->i_gid = dir->i_gid;
		if (S_ISDIR(mode))
			mode |= S_ISGID;
	} else
		inode->i_gid = current_fsgid();

	inode->i_mode = mode;
	inode->i_ino = objectid;
	inode_set_bytes(inode, 0);
	inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME;
	inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0],
				  struct btrfs_inode_item);
	fill_inode_item(trans, path->nodes[0], inode_item, inode);

	ref = btrfs_item_ptr(path->nodes[0], path->slots[0] + 1,
			     struct btrfs_inode_ref);
	btrfs_set_inode_ref_name_len(path->nodes[0], ref, name_len);
	btrfs_set_inode_ref_index(path->nodes[0], ref, *index);
	ptr = (unsigned long)(ref + 1);
	write_extent_buffer(path->nodes[0], name, ptr, name_len);

	btrfs_mark_buffer_dirty(path->nodes[0]);
	btrfs_free_path(path);

	location = &BTRFS_I(inode)->location;
	location->objectid = objectid;
	location->offset = 0;
	btrfs_set_key_type(location, BTRFS_INODE_ITEM_KEY);

	btrfs_inherit_iflags(inode, dir);

	if ((mode & S_IFREG)) {
		if (btrfs_test_opt(root, NODATASUM))
			BTRFS_I(inode)->flags |= BTRFS_INODE_NODATASUM;
		if (btrfs_test_opt(root, NODATACOW))
			BTRFS_I(inode)->flags |= BTRFS_INODE_NODATACOW;
	}

	insert_inode_hash(inode);
	inode_tree_add(inode);
	return inode;
fail:
	if (dir)
		BTRFS_I(dir)->index_cnt--;
	btrfs_free_path(path);
	iput(inode);
	return ERR_PTR(ret);
}

static inline u8 btrfs_inode_type(struct inode *inode)
{
	return btrfs_type_by_mode[(inode->i_mode & S_IFMT) >> S_SHIFT];
}

/*
 * utility function to add 'inode' into 'parent_inode' with
 * a give name and a given sequence number.
 * if 'add_backref' is true, also insert a backref from the
 * inode to the parent directory.
 */
int btrfs_add_link(struct btrfs_trans_handle *trans,
		   struct inode *parent_inode, struct inode *inode,
		   const char *name, int name_len, int add_backref, u64 index)
{
	int ret;
	struct btrfs_key key;
	struct btrfs_root *root = BTRFS_I(parent_inode)->root;

	key.objectid = inode->i_ino;
	btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY);
	key.offset = 0;

	ret = btrfs_insert_dir_item(trans, root, name, name_len,
				    parent_inode->i_ino,
				    &key, btrfs_inode_type(inode),
				    index);
	if (ret == 0) {
		if (add_backref) {
			ret = btrfs_insert_inode_ref(trans, root,
						     name, name_len,
						     inode->i_ino,
						     parent_inode->i_ino,
						     index);
		}
		btrfs_i_size_write(parent_inode, parent_inode->i_size +
				   name_len * 2);
		parent_inode->i_mtime = parent_inode->i_ctime = CURRENT_TIME;
		ret = btrfs_update_inode(trans, root, parent_inode);
	}
	return ret;
}

static int btrfs_add_nondir(struct btrfs_trans_handle *trans,
			    struct dentry *dentry, struct inode *inode,
			    int backref, u64 index)
{
	int err = btrfs_add_link(trans, dentry->d_parent->d_inode,
				 inode, dentry->d_name.name,
				 dentry->d_name.len, backref, index);
	if (!err) {
		d_instantiate(dentry, inode);
		return 0;
	}
	if (err > 0)
		err = -EEXIST;
	return err;
}

static int btrfs_mknod(struct inode *dir, struct dentry *dentry,
			int mode, dev_t rdev)
{
	struct btrfs_trans_handle *trans;
	struct btrfs_root *root = BTRFS_I(dir)->root;
	struct inode *inode = NULL;
	int err;
	int drop_inode = 0;
	u64 objectid;
	unsigned long nr = 0;
	u64 index = 0;

	if (!new_valid_dev(rdev))
		return -EINVAL;

	err = btrfs_check_metadata_free_space(root);
	if (err)
		goto fail;

	trans = btrfs_start_transaction(root, 1);
	btrfs_set_trans_block_group(trans, dir);

	err = btrfs_find_free_objectid(trans, root, dir->i_ino, &objectid);
	if (err) {
		err = -ENOSPC;
		goto out_unlock;
	}

	inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
				dentry->d_name.len,
				dentry->d_parent->d_inode->i_ino, objectid,
				BTRFS_I(dir)->block_group, mode, &index);
	err = PTR_ERR(inode);
	if (IS_ERR(inode))
		goto out_unlock;

	err = btrfs_init_inode_security(inode, dir);
	if (err) {
		drop_inode = 1;
		goto out_unlock;
	}

	btrfs_set_trans_block_group(trans, inode);
	err = btrfs_add_nondir(trans, dentry, inode, 0, index);
	if (err)
		drop_inode = 1;
	else {
		inode->i_op = &btrfs_special_inode_operations;
		init_special_inode(inode, inode->i_mode, rdev);
		btrfs_update_inode(trans, root, inode);
	}
	btrfs_update_inode_block_group(trans, inode);
	btrfs_update_inode_block_group(trans, dir);
out_unlock:
	nr = trans->blocks_used;
	btrfs_end_transaction_throttle(trans, root);
fail:
	if (drop_inode) {
		inode_dec_link_count(inode);
		iput(inode);
	}
	btrfs_btree_balance_dirty(root, nr);
	return err;
}

static int btrfs_create(struct inode *dir, struct dentry *dentry,
			int mode, struct nameidata *nd)
{
	struct btrfs_trans_handle *trans;
	struct btrfs_root *root = BTRFS_I(dir)->root;
	struct inode *inode = NULL;
	int err;
	int drop_inode = 0;
	unsigned long nr = 0;
	u64 objectid;
	u64 index = 0;

	err = btrfs_check_metadata_free_space(root);
	if (err)
		goto fail;
	trans = btrfs_start_transaction(root, 1);
	btrfs_set_trans_block_group(trans, dir);

	err = btrfs_find_free_objectid(trans, root, dir->i_ino, &objectid);
	if (err) {
		err = -ENOSPC;
		goto out_unlock;
	}

	inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
				dentry->d_name.len,
				dentry->d_parent->d_inode->i_ino,
				objectid, BTRFS_I(dir)->block_group, mode,
				&index);
	err = PTR_ERR(inode);
	if (IS_ERR(inode))
		goto out_unlock;

	err = btrfs_init_inode_security(inode, dir);
	if (err) {
		drop_inode = 1;
		goto out_unlock;
	}

	btrfs_set_trans_block_group(trans, inode);
	err = btrfs_add_nondir(trans, dentry, inode, 0, index);
	if (err)
		drop_inode = 1;
	else {
		inode->i_mapping->a_ops = &btrfs_aops;
		inode->i_mapping->backing_dev_info = &root->fs_info->bdi;
		inode->i_fop = &btrfs_file_operations;
		inode->i_op = &btrfs_file_inode_operations;
		BTRFS_I(inode)->io_tree.ops = &btrfs_extent_io_ops;
	}
	btrfs_update_inode_block_group(trans, inode);
	btrfs_update_inode_block_group(trans, dir);
out_unlock:
	nr = trans->blocks_used;
	btrfs_end_transaction_throttle(trans, root);
fail:
	if (drop_inode) {
		inode_dec_link_count(inode);
		iput(inode);
	}
	btrfs_btree_balance_dirty(root, nr);
	return err;
}

static int btrfs_link(struct dentry *old_dentry, struct inode *dir,
		      struct dentry *dentry)
{
	struct btrfs_trans_handle *trans;
	struct btrfs_root *root = BTRFS_I(dir)->root;
	struct inode *inode = old_dentry->d_inode;
	u64 index;
	unsigned long nr = 0;
	int err;
	int drop_inode = 0;

	if (inode->i_nlink == 0)
		return -ENOENT;

	btrfs_inc_nlink(inode);
	err = btrfs_check_metadata_free_space(root);
	if (err)
		goto fail;
	err = btrfs_set_inode_index(dir, &index);
	if (err)
		goto fail;

	trans = btrfs_start_transaction(root, 1);

	btrfs_set_trans_block_group(trans, dir);
	atomic_inc(&inode->i_count);

	err = btrfs_add_nondir(trans, dentry, inode, 1, index);

	if (err)
		drop_inode = 1;

	btrfs_update_inode_block_group(trans, dir);
	err = btrfs_update_inode(trans, root, inode);

	if (err)
		drop_inode = 1;

	nr = trans->blocks_used;

	btrfs_log_new_name(trans, inode, NULL, dentry->d_parent);
	btrfs_end_transaction_throttle(trans, root);
fail:
	if (drop_inode) {
		inode_dec_link_count(inode);
		iput(inode);
	}
	btrfs_btree_balance_dirty(root, nr);
	return err;
}

static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode)
{
	struct inode *inode = NULL;
	struct btrfs_trans_handle *trans;
	struct btrfs_root *root = BTRFS_I(dir)->root;
	int err = 0;
	int drop_on_err = 0;
	u64 objectid = 0;
	u64 index = 0;
	unsigned long nr = 1;

	err = btrfs_check_metadata_free_space(root);
	if (err)
		goto out_unlock;

	trans = btrfs_start_transaction(root, 1);
	btrfs_set_trans_block_group(trans, dir);

	if (IS_ERR(trans)) {
		err = PTR_ERR(trans);
		goto out_unlock;
	}

	err = btrfs_find_free_objectid(trans, root, dir->i_ino, &objectid);
	if (err) {
		err = -ENOSPC;
		goto out_unlock;
	}

	inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name,
				dentry->d_name.len,
				dentry->d_parent->d_inode->i_ino, objectid,
				BTRFS_I(dir)->block_group, S_IFDIR | mode,
				&index);
	if (IS_ERR(inode)) {
		err = PTR_ERR(inode);
		goto out_fail;
	}

	drop_on_err = 1;

	err = btrfs_init_inode_security(inode, dir);
	if (err)
		goto out_fail;

	inode->i_op = &btrfs_dir_inode_operations;
	inode->i_fop = &btrfs_dir_file_operations;
	btrfs_set_trans_block_group(trans, inode);

	btrfs_i_size_write(inode, 0);
	err = btrfs_update_inode(trans, root, inode);
	if (err)
		goto out_fail;

	err = btrfs_add_link(trans, dentry->d_parent->d_inode,
				 inode, dentry->d_name.name,
				 dentry->d_name.len, 0, index);
	if (err)
		goto out_fail;

	d_instantiate(dentry, inode);
	drop_on_err = 0;
	btrfs_update_inode_block_group(trans, inode);
	btrfs_update_inode_block_group(trans, dir);

out_fail:
	nr = trans->blocks_used;
	btrfs_end_transaction_throttle(trans, root);

out_unlock:
	if (drop_on_err)
		iput(inode);
	btrfs_btree_balance_dirty(root, nr);
	return err;
}

/* helper for btfs_get_extent.  Given an existing extent in the tree,
 * and an extent that you want to insert, deal with overlap and insert
 * the new extent into the tree.
 */
static int merge_extent_mapping(struct extent_map_tree *em_tree,
				struct extent_map *existing,
				struct extent_map *em,
				u64 map_start, u64 map_len)
{
	u64 start_diff;

	BUG_ON(map_start < em->start || map_start >= extent_map_end(em));
	start_diff = map_start - em->start;
	em->start = map_start;
	em->len = map_len;
	if (em->block_start < EXTENT_MAP_LAST_BYTE &&
	    !test_bit(EXTENT_FLAG_COMPRESSED, &em->flags)) {
		em->block_start += start_diff;
		em->block_len -= start_diff;
	}
	return add_extent_mapping(em_tree, em);
}

static noinline int uncompress_inline(struct btrfs_path *path,
				      struct inode *inode, struct page *page,
				      size_t pg_offset, u64 extent_offset,
				      struct btrfs_file_extent_item *item)
{
	int ret;
	struct extent_buffer *leaf = path->nodes[0];
	char *tmp;
	size_t max_size;
	unsigned long inline_size;
	unsigned long ptr;

	WARN_ON(pg_offset != 0);
	max_size = btrfs_file_extent_ram_bytes(leaf, item);
	inline_size = btrfs_file_extent_inline_item_len(leaf,
					btrfs_item_nr(leaf, path->slots[0]));
	tmp = kmalloc(inline_size, GFP_NOFS);
	ptr = btrfs_file_extent_inline_start(item);

	read_extent_buffer(leaf, tmp, ptr, inline_size);

	max_size = min_t(unsigned long, PAGE_CACHE_SIZE, max_size);
	ret = btrfs_zlib_decompress(tmp, page, extent_offset,
				    inline_size, max_size);
	if (ret) {
		char *kaddr = kmap_atomic(page, KM_USER0);
		unsigned long copy_size = min_t(u64,
				  PAGE_CACHE_SIZE - pg_offset,
				  max_size - extent_offset);
		memset(kaddr + pg_offset, 0, copy_size);
		kunmap_atomic(kaddr, KM_USER0);
	}
	kfree(tmp);
	return 0;
}

/*
 * a bit scary, this does extent mapping from logical file offset to the disk.
 * the ugly parts come from merging extents from the disk with the in-ram
 * representation.  This gets more complex because of the data=ordered code,
 * where the in-ram extents might be locked pending data=ordered completion.
 *
 * This also copies inline extents directly into the page.
 */

struct extent_map *btrfs_get_extent(struct inode *inode, struct page *page,
				    size_t pg_offset, u64 start, u64 len,
				    int create)
{
	int ret;
	int err = 0;
	u64 bytenr;
	u64 extent_start = 0;
	u64 extent_end = 0;
	u64 objectid = inode->i_ino;
	u32 found_type;
	struct btrfs_path *path = NULL;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct btrfs_file_extent_item *item;
	struct extent_buffer *leaf;
	struct btrfs_key found_key;
	struct extent_map *em = NULL;
	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct btrfs_trans_handle *trans = NULL;
	int compressed;

again:
	spin_lock(&em_tree->lock);
	em = lookup_extent_mapping(em_tree, start, len);
	if (em)
		em->bdev = root->fs_info->fs_devices->latest_bdev;
	spin_unlock(&em_tree->lock);

	if (em) {
		if (em->start > start || em->start + em->len <= start)
			free_extent_map(em);
		else if (em->block_start == EXTENT_MAP_INLINE && page)
			free_extent_map(em);
		else
			goto out;
	}
	em = alloc_extent_map(GFP_NOFS);
	if (!em) {
		err = -ENOMEM;
		goto out;
	}
	em->bdev = root->fs_info->fs_devices->latest_bdev;
	em->start = EXTENT_MAP_HOLE;
	em->orig_start = EXTENT_MAP_HOLE;
	em->len = (u64)-1;
	em->block_len = (u64)-1;

	if (!path) {
		path = btrfs_alloc_path();
		BUG_ON(!path);
	}

	ret = btrfs_lookup_file_extent(trans, root, path,
				       objectid, start, trans != NULL);
	if (ret < 0) {
		err = ret;
		goto out;
	}

	if (ret != 0) {
		if (path->slots[0] == 0)
			goto not_found;
		path->slots[0]--;
	}

	leaf = path->nodes[0];
	item = btrfs_item_ptr(leaf, path->slots[0],
			      struct btrfs_file_extent_item);
	/* are we inside the extent that was found? */
	btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
	found_type = btrfs_key_type(&found_key);
	if (found_key.objectid != objectid ||
	    found_type != BTRFS_EXTENT_DATA_KEY) {
		goto not_found;
	}

	found_type = btrfs_file_extent_type(leaf, item);
	extent_start = found_key.offset;
	compressed = btrfs_file_extent_compression(leaf, item);
	if (found_type == BTRFS_FILE_EXTENT_REG ||
	    found_type == BTRFS_FILE_EXTENT_PREALLOC) {
		extent_end = extent_start +
		       btrfs_file_extent_num_bytes(leaf, item);
	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
		size_t size;
		size = btrfs_file_extent_inline_len(leaf, item);
		extent_end = (extent_start + size + root->sectorsize - 1) &
			~((u64)root->sectorsize - 1);
	}

	if (start >= extent_end) {
		path->slots[0]++;
		if (path->slots[0] >= btrfs_header_nritems(leaf)) {
			ret = btrfs_next_leaf(root, path);
			if (ret < 0) {
				err = ret;
				goto out;
			}
			if (ret > 0)
				goto not_found;
			leaf = path->nodes[0];
		}
		btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]);
		if (found_key.objectid != objectid ||
		    found_key.type != BTRFS_EXTENT_DATA_KEY)
			goto not_found;
		if (start + len <= found_key.offset)
			goto not_found;
		em->start = start;
		em->len = found_key.offset - start;
		goto not_found_em;
	}

	if (found_type == BTRFS_FILE_EXTENT_REG ||
	    found_type == BTRFS_FILE_EXTENT_PREALLOC) {
		em->start = extent_start;
		em->len = extent_end - extent_start;
		em->orig_start = extent_start -
				 btrfs_file_extent_offset(leaf, item);
		bytenr = btrfs_file_extent_disk_bytenr(leaf, item);
		if (bytenr == 0) {
			em->block_start = EXTENT_MAP_HOLE;
			goto insert;
		}
		if (compressed) {
			set_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
			em->block_start = bytenr;
			em->block_len = btrfs_file_extent_disk_num_bytes(leaf,
									 item);
		} else {
			bytenr += btrfs_file_extent_offset(leaf, item);
			em->block_start = bytenr;
			em->block_len = em->len;
			if (found_type == BTRFS_FILE_EXTENT_PREALLOC)
				set_bit(EXTENT_FLAG_PREALLOC, &em->flags);
		}
		goto insert;
	} else if (found_type == BTRFS_FILE_EXTENT_INLINE) {
		unsigned long ptr;
		char *map;
		size_t size;
		size_t extent_offset;
		size_t copy_size;

		em->block_start = EXTENT_MAP_INLINE;
		if (!page || create) {
			em->start = extent_start;
			em->len = extent_end - extent_start;
			goto out;
		}

		size = btrfs_file_extent_inline_len(leaf, item);
		extent_offset = page_offset(page) + pg_offset - extent_start;
		copy_size = min_t(u64, PAGE_CACHE_SIZE - pg_offset,
				size - extent_offset);
		em->start = extent_start + extent_offset;
		em->len = (copy_size + root->sectorsize - 1) &
			~((u64)root->sectorsize - 1);
		em->orig_start = EXTENT_MAP_INLINE;
		if (compressed)
			set_bit(EXTENT_FLAG_COMPRESSED, &em->flags);
		ptr = btrfs_file_extent_inline_start(item) + extent_offset;
		if (create == 0 && !PageUptodate(page)) {
			if (btrfs_file_extent_compression(leaf, item) ==
			    BTRFS_COMPRESS_ZLIB) {
				ret = uncompress_inline(path, inode, page,
							pg_offset,
							extent_offset, item);
				BUG_ON(ret);
			} else {
				map = kmap(page);
				read_extent_buffer(leaf, map + pg_offset, ptr,
						   copy_size);
				kunmap(page);
			}
			flush_dcache_page(page);
		} else if (create && PageUptodate(page)) {
			if (!trans) {
				kunmap(page);
				free_extent_map(em);
				em = NULL;
				btrfs_release_path(root, path);
				trans = btrfs_join_transaction(root, 1);
				goto again;
			}
			map = kmap(page);
			write_extent_buffer(leaf, map + pg_offset, ptr,
					    copy_size);
			kunmap(page);
			btrfs_mark_buffer_dirty(leaf);
		}
		set_extent_uptodate(io_tree, em->start,
				    extent_map_end(em) - 1, GFP_NOFS);
		goto insert;
	} else {
		printk(KERN_ERR "btrfs unknown found_type %d\n", found_type);
		WARN_ON(1);
	}
not_found:
	em->start = start;
	em->len = len;
not_found_em:
	em->block_start = EXTENT_MAP_HOLE;
	set_bit(EXTENT_FLAG_VACANCY, &em->flags);
insert:
	btrfs_release_path(root, path);
	if (em->start > start || extent_map_end(em) <= start) {
		printk(KERN_ERR "Btrfs: bad extent! em: [%llu %llu] passed "
		       "[%llu %llu]\n", (unsigned long long)em->start,
		       (unsigned long long)em->len,
		       (unsigned long long)start,
		       (unsigned long long)len);
		err = -EIO;
		goto out;
	}

	err = 0;
	spin_lock(&em_tree->lock);
	ret = add_extent_mapping(em_tree, em);
	/* it is possible that someone inserted the extent into the tree
	 * while we had the lock dropped.  It is also possible that
	 * an overlapping map exists in the tree
	 */
	if (ret == -EEXIST) {
		struct extent_map *existing;

		ret = 0;

		existing = lookup_extent_mapping(em_tree, start, len);
		if (existing && (existing->start > start ||
		    existing->start + existing->len <= start)) {
			free_extent_map(existing);
			existing = NULL;
		}
		if (!existing) {
			existing = lookup_extent_mapping(em_tree, em->start,
							 em->len);
			if (existing) {
				err = merge_extent_mapping(em_tree, existing,
							   em, start,
							   root->sectorsize);
				free_extent_map(existing);
				if (err) {
					free_extent_map(em);
					em = NULL;
				}
			} else {
				err = -EIO;
				free_extent_map(em);
				em = NULL;
			}
		} else {
			free_extent_map(em);
			em = existing;
			err = 0;
		}
	}
	spin_unlock(&em_tree->lock);
out:
	if (path)
		btrfs_free_path(path);
	if (trans) {
		ret = btrfs_end_transaction(trans, root);
		if (!err)
			err = ret;
	}
	if (err) {
		free_extent_map(em);
		return ERR_PTR(err);
	}
	return em;
}

static ssize_t btrfs_direct_IO(int rw, struct kiocb *iocb,
			const struct iovec *iov, loff_t offset,
			unsigned long nr_segs)
{
	return -EINVAL;
}

static int btrfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
		__u64 start, __u64 len)
{
	return extent_fiemap(inode, fieinfo, start, len, btrfs_get_extent);
}

int btrfs_readpage(struct file *file, struct page *page)
{
	struct extent_io_tree *tree;
	tree = &BTRFS_I(page->mapping->host)->io_tree;
	return extent_read_full_page(tree, page, btrfs_get_extent);
}

static int btrfs_writepage(struct page *page, struct writeback_control *wbc)
{
	struct extent_io_tree *tree;


	if (current->flags & PF_MEMALLOC) {
		redirty_page_for_writepage(wbc, page);
		unlock_page(page);
		return 0;
	}
	tree = &BTRFS_I(page->mapping->host)->io_tree;
	return extent_write_full_page(tree, page, btrfs_get_extent, wbc);
}

int btrfs_writepages(struct address_space *mapping,
		     struct writeback_control *wbc)
{
	struct extent_io_tree *tree;

	tree = &BTRFS_I(mapping->host)->io_tree;
	return extent_writepages(tree, mapping, btrfs_get_extent, wbc);
}

static int
btrfs_readpages(struct file *file, struct address_space *mapping,
		struct list_head *pages, unsigned nr_pages)
{
	struct extent_io_tree *tree;
	tree = &BTRFS_I(mapping->host)->io_tree;
	return extent_readpages(tree, mapping, pages, nr_pages,
				btrfs_get_extent);
}
static int __btrfs_releasepage(struct page *page, gfp_t gfp_flags)
{
	struct extent_io_tree *tree;
	struct extent_map_tree *map;
	int ret;

	tree = &BTRFS_I(page->mapping->host)->io_tree;
	map = &BTRFS_I(page->mapping->host)->extent_tree;
	ret = try_release_extent_mapping(map, tree, page, gfp_flags);
	if (ret == 1) {
		ClearPagePrivate(page);
		set_page_private(page, 0);
		page_cache_release(page);
	}
	return ret;
}

static int btrfs_releasepage(struct page *page, gfp_t gfp_flags)
{
	if (PageWriteback(page) || PageDirty(page))
		return 0;
	return __btrfs_releasepage(page, gfp_flags & GFP_NOFS);
}

static void btrfs_invalidatepage(struct page *page, unsigned long offset)
{
	struct extent_io_tree *tree;
	struct btrfs_ordered_extent *ordered;
	u64 page_start = page_offset(page);
	u64 page_end = page_start + PAGE_CACHE_SIZE - 1;

	wait_on_page_writeback(page);
	tree = &BTRFS_I(page->mapping->host)->io_tree;
	if (offset) {
		btrfs_releasepage(page, GFP_NOFS);
		return;
	}

	lock_extent(tree, page_start, page_end, GFP_NOFS);
	ordered = btrfs_lookup_ordered_extent(page->mapping->host,
					   page_offset(page));
	if (ordered) {
		/*
		 * IO on this page will never be started, so we need
		 * to account for any ordered extents now
		 */
		clear_extent_bit(tree, page_start, page_end,
				 EXTENT_DIRTY | EXTENT_DELALLOC |
				 EXTENT_LOCKED, 1, 0, GFP_NOFS);
		btrfs_finish_ordered_io(page->mapping->host,
					page_start, page_end);
		btrfs_put_ordered_extent(ordered);
		lock_extent(tree, page_start, page_end, GFP_NOFS);
	}
	clear_extent_bit(tree, page_start, page_end,
		 EXTENT_LOCKED | EXTENT_DIRTY | EXTENT_DELALLOC |
		 EXTENT_ORDERED,
		 1, 1, GFP_NOFS);
	__btrfs_releasepage(page, GFP_NOFS);

	ClearPageChecked(page);
	if (PagePrivate(page)) {
		ClearPagePrivate(page);
		set_page_private(page, 0);
		page_cache_release(page);
	}
}

/*
 * btrfs_page_mkwrite() is not allowed to change the file size as it gets
 * called from a page fault handler when a page is first dirtied. Hence we must
 * be careful to check for EOF conditions here. We set the page up correctly
 * for a written page which means we get ENOSPC checking when writing into
 * holes and correct delalloc and unwritten extent mapping on filesystems that
 * support these features.
 *
 * We are not allowed to take the i_mutex here so we have to play games to
 * protect against truncate races as the page could now be beyond EOF.  Because
 * vmtruncate() writes the inode size before removing pages, once we have the
 * page lock we can determine safely if the page is beyond EOF. If it is not
 * beyond EOF, then the page is guaranteed safe against truncation until we
 * unlock the page.
 */
int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf)
{
	struct page *page = vmf->page;
	struct inode *inode = fdentry(vma->vm_file)->d_inode;
	struct btrfs_root *root = BTRFS_I(inode)->root;
	struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree;
	struct btrfs_ordered_extent *ordered;
	char *kaddr;
	unsigned long zero_start;
	loff_t size;
	int ret;
	u64 page_start;
	u64 page_end;

	ret = btrfs_check_data_free_space(root, inode, PAGE_CACHE_SIZE);
	if (ret) {
		if (ret == -ENOMEM)
			ret = VM_FAULT_OOM;
		else /* -ENOSPC, -EIO, etc */
			ret = VM_FAULT_SIGBUS;
		goto out;
	}

	ret = VM_FAULT_NOPAGE; /* make the VM retry the fault */
again:
	lock_page(page);
	size = i_size_read(inode);
	page_start = page_offset(page);
	page_end = page_start + PAGE_CACHE_SIZE - 1;

	if ((page->mapping != inode->i_mapping) ||
	    (page_start >= size)) {
		btrfs_free_reserved_data_space(root, inode, PAGE_CACHE_SIZE);
		/* page got truncated out from underneath us */
		goto out_unlock;
	}
	wait_on_page_writeback(page);

	lock_extent(io_tree, page_start, page_end, GFP_NOFS);
	set_page_extent_mapped(page);

	/*
	 * we can't set the delalloc bits if there are pending ordered
	 * extents.  Drop our locks and wait for them to finish
	 */
	ordered = btrfs_lookup_ordered_extent(inode, page_start);
	if (ordered) {
		unlock_extent(io_tree, page_start, page_end, GFP_NOFS);
		unlock_page(page);
		btrfs_start_ordered_extent(inode, ordered, 1);
		btrfs_put_ordered_extent(ordered);
		goto again;
	}

	btrfs_set_extent_delalloc(inode, page_start, page_end);
	ret = 0;

	/* page is wholly or partially inside EOF */
	if (page_start + PAGE_CACHE_SIZE > size)
		zero_start = size & ~PAGE_CACHE_MASK;
	else
		zero_start = PAGE_CACHE_SIZE;

	if (zero_start != PAGE_CACHE_SIZE) {
		kaddr = kmap(page);
		memset(kaddr + zero_start, 0, PAGE_CACHE_SIZE - zero_start);
		flush_dcache_page(page);
		kunmap(page);
	}
	ClearPageChecked(page);
	set_page_dirty(page);

	BTRFS_I(inode)->last_trans = root->fs_info->generation + 1;
	unlock_extent(io_tree, page_start, page_end, GFP_NOFS);

out_unlock:
	unlock_page(page);
out:
	return ret;
}

static void btrfs_truncate(struct inode *inode)
{
	struct btrfs_root *root = BTRFS_I(inode)->root;
	int ret;
	struct btrfs_trans_handle *trans;
	unsigned long nr;
	u64 mask = root->sectorsize - 1;

	if (!S_ISREG(inode->i_mode))
		return;
	if (IS_APPEND(inode) || IS_IMMUTABLE(inode))
		return;

	btrfs_truncate_page(inode->i_mapping, inode->i_size);
	btrfs_wait_ordered_range(inode, inode->i_size & (~mask), (u64)-1);

	trans = btrfs_start_transaction(root, 1);

	/*
	 * setattr is responsible for setting the ordered_data_close flag,
	 * but that is only tested during the last file release.  That
	 * could happen well after the next commit, leaving a great big
	 * window where new writes may get lost if someone chooses to write
	 * to this file after truncating to zero
	 *
	 * The inode doesn't have any dirty data here, and so if we commit
	 * this is a noop.  If someone immediately starts writing to the inode
	 * it is very likely we'll catch some of their writes in this
	 * transaction, and the commit will find this file on the ordered
	 * data list with good things to send down.
	 *
	 * This is a best effort solution, there is still a window where
	 * using truncate to replace the contents of the file will
	 * end up with a zero length file after a crash.
	 */
	if (inode->i_size == 0 && BTRFS_I(inode)->ordered_data_close)
		btrfs_add_ordered_operation(trans, root, inode);

	btrfs_set_trans_block_group(trans, inode);
	btrfs_i_size_write(inode, inode->i_size);

	ret = btrfs_orphan_add(trans, inode);
	if (ret)
		goto out;
	/* FIXME, add redo link to tree so we don't leak on crash */
	ret = btrfs_truncate_inode_items(trans, root, inode, inode->i_size,
				      BTRFS_EXTENT_DATA_KEY);
	btrfs_update_inode(trans, root, inode);

	ret = btrfs_orphan_del(trans, inode);
	BUG_ON(ret);

out:
	nr = trans->blocks_used;
	ret = btrfs_end_transaction_throttle(trans, root);
	BUG_ON(ret);
	btrfs_btree_balance_dirty(root, nr);
}

/*
 * create a new subvolume directory/inode (helper for the ioctl).
 */
int btrfs_create_subvol_root(struct btrfs_trans_handle *trans,
			     struct btrfs_root *new_root, struct dentry *dentry,
			     u64 new_dirid, u64 alloc_hint)
{
	struct inode *inode;
	int error;
	u64 index = 0;

	inode = btrfs_new_inode(trans, new_root, NULL, "..", 2, new_dirid,
				new_dirid, alloc_hint, S_IFDIR | 0700, &index);
	if (IS_ERR(inode))
		return PTR_ERR(inode);
	inode->i_op = &btrfs_dir_inode_operations;
	inode->i_fop = &btrfs_dir_file_operations;

	inode->i_nlink = 1;
	btrfs_i_size_write(inode, 0);

	error = btrfs_update_inode(trans, new_root, inode);
	if (error)
		return error;

	d_instantiate(dentry, inode);
	return 0;
}

/* helper function for file defrag and space balancing.  This
 * forces readahead on a given range of bytes in an inode
 */
unsigned long btrfs_force_ra(struct address_space *mapping,
			      struct file_ra_state *ra, struct file *file,
			      pgoff_t offset, pgoff_t last_index)
{
	pgoff_t req_size = last_index - offset + 1;

	page_cache_sync_readahead(mapping, ra, file, offset, req_size);
	return offset + req_size;
}

struct inode *btrfs_alloc_inode(struct super_block *sb)
{
	struct btrfs_inode *ei;

	ei = kmem_cache_alloc(btrfs_inode_cachep, GFP_NOFS);
	if (!ei)
		return NULL;
	ei->last_trans = 0;
	ei->logged_trans = 0;
	btrfs_ordered_inode_tree_init(&ei->ordered_tree);
	INIT_LIST_HEAD(&ei->i_orphan);
	INIT_LIST_HEAD(&ei->ordered_operations);
	return &ei->vfs_inode;
}

void btrfs_destroy_inode(struct inode *inode)
{
	struct btrfs_ordered_extent *ordered;
	struct btrfs_root *root = BTRFS_I(inode)->root;

	WARN_ON(!list_empty(&inode->i_dentry));
	WARN_ON(inode->i_data.nrpages);

	/*
	 * Make sure we're properly removed from the ordered operation
	 * lists.
	 */
	smp_mb();
	if (!list_empty(&BTRFS_I(inode)->ordered_operations)) {
		spin_lock(&root->fs_info->ordered_extent_lock);
		list_del_init(&BTRFS_I(inode)->ordered_operations);
		spin_unlock(&root->fs_info->ordered_extent_lock);
	}

	spin_lock(&root->list_lock);
	if (!list_empty(&BTRFS_I(inode)->i_orphan)) {
		printk(KERN_ERR "BTRFS: inode %lu: inode still on the orphan"
		       " list\n", inode->i_ino);
		dump_stack();
	}
	spin_unlock(&root->list_lock);

	while (1) {
		ordered = btrfs_lookup_first_ordered_extent(inode, (u64)-1);
		if (!ordered)
			break;
		else {
			printk(KERN_ERR "btrfs found ordered "
			       "extent %llu %llu on inode cleanup\n",
			       (unsigned long long)ordered->file_offset,
			       (unsigned long long)ordered->len);
			btrfs_remove_ordered_extent(inode, ordered);
			btrfs_put_ordered_extent(ordered);
			btrfs_put_ordered_extent(ordered);
		}
	}
	inode_tree_del(inode);
	btrfs_drop_extent_cache(inode, 0, (u64)-1, 0);
	kmem_cache_free(btrfs_inode_cachep, BTRFS_I(inode));
}

static void init_once(void *foo)
{
	struct btrfs_inode *ei = (struct btrfs_inode *) foo;

	inode_init_once(&ei->vfs_inode);
}

void btrfs_destroy_cachep(void)
{
	if (btrfs_inode_cachep)
		kmem_cache_destroy(btrfs_inode_cachep);
	if (btrfs_trans_handle_cachep)
		kmem_cache_destroy(btrfs_trans_handle_cachep);
	if (btrfs_transaction_cachep)
		kmem_cache_destroy(btrfs_transaction_cachep);
	if (btrfs_path_cachep)
		kmem_cache_destroy(btrfs_path_cachep);
}

int btrfs_init_cachep(void)
{
	btrfs_inode_cachep = kmem_cache_create("btrfs_inode_cache",
			sizeof(struct btrfs_inode), 0,
			SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, init_once);
	if (!btrfs_inode_cachep)
		goto fail;

	btrfs_trans_handle_cachep = kmem_cache_create("btrfs_trans_handle_cache",
			sizeof(struct btrfs_trans_handle), 0,
			SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, NULL);
	if (!btrfs_trans_handle_cachep)
		goto fail;

	btrfs_transaction_cachep = kmem_cache_create("btrfs_transaction_cache",
			sizeof(struct btrfs_transaction), 0,
			SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, NULL);
	if (!btrfs_transaction_cachep)
		goto fail;

	btrfs_path_cachep = kmem_cache_create("btrfs_path_cache",
			sizeof(struct btrfs_path), 0,
			SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, NULL);
	if (!btrfs_path_cachep)
		goto fail;

	return 0;
fail:
	btrfs_destroy_cachep();
	return -ENOMEM;
}

static int btrfs_getattr(struct vfsmount *mnt,
			 struct dentry *dentry, struct kstat *stat)
{
	struct inode *inode = dentry->d_inode;
	generic_fillattr(inode, stat);
	stat->dev = BTRFS_I(inode)->root->anon_super.s_dev;
	stat->blksize = PAGE_CACHE_SIZE;
	stat->blocks = (inode_get_bytes(inode) +
			BTRFS_I(inode)->delalloc_bytes) >> 9;
	return 0;
}

static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry,
			   struct inode *new_dir, struct dentry *new_dentry)
{
	struct btrfs_trans_handle *trans;
	struct btrfs_root *root = BTRFS_I(old_dir)->root;
	struct inode *new_inode = new_dentry->d_inode;
	struct inode *old_inode = old_dentry->d_inode;
	struct timespec ctime = CURRENT_TIME;
	u64 index = 0;
	int ret;

	/* we're not allowed to rename between subvolumes */
	if (BTRFS_I(old_inode)->root->root_key.objectid !=
	    BTRFS_I(new_dir)->root->root_key.objectid)
		return -EXDEV;

	if (S_ISDIR(old_inode->i_mode) && new_inode &&
	    new_inode->i_size > BTRFS_EMPTY_DIR_SIZE) {
		return -ENOTEMPTY;
	}

	/* to rename a snapshot or subvolume, we need to juggle the
	 * backrefs.  This isn't coded yet
	 */
	if (old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)
		return -EXDEV;

	ret = btrfs_check_metadata_free_space(root);
	if (ret)
		goto out_unlock;

	/*
	 * we're using rename to replace one file with another.
	 * and the replacement file is large.  Start IO on it now so
	 * we don't add too much work to the end of the transaction
	 */
	if (new_inode && S_ISREG(old_inode->i_mode) && new_inode->i_size &&
	    old_inode->i_size > BTRFS_ORDERED_OPERATIONS_FLUSH_LIMIT)
		filemap_flush(old_inode->i_mapping);

	trans = btrfs_start_transaction(root, 1);

	/*
	 * make sure the inode gets flushed if it is replacing
	 * something.
	 */
	if (new_inode && new_inode->i_size &&
	    old_inode && S_ISREG(old_inode->i_mode)) {
		btrfs_add_ordered_operation(trans, root, old_inode);
	}

	/*
	 * this is an ugly little race, but the rename is required to make
	 * sure that if we crash, the inode is either at the old name
	 * or the new one.  pinning the log transaction lets us make sure
	 * we don't allow a log commit to come in after we unlink the
	 * name but before we add the new name back in.
	 */
	btrfs_pin_log_trans(root);

	btrfs_set_trans_block_group(trans, new_dir);

	btrfs_inc_nlink(old_dentry->d_inode);
	old_dir->i_ctime = old_dir->i_mtime = ctime;
	new_dir->i_ctime = new_dir->i_mtime = ctime;
	old_inode->i_ctime = ctime;

	if (old_dentry->d_parent != new_dentry->d_parent)
		btrfs_record_unlink_dir(trans, old_dir, old_inode, 1);

	ret = btrfs_unlink_inode(trans, root, old_dir, old_dentry->d_inode,
				 old_dentry->d_name.name,
				 old_dentry->d_name.len);
	if (ret)
		goto out_fail;

	if (new_inode) {
		new_inode->i_ctime = CURRENT_TIME;
		ret = btrfs_unlink_inode(trans, root, new_dir,
					 new_dentry->d_inode,
					 new_dentry->d_name.name,
					 new_dentry->d_name.len);
		if (ret)
			goto out_fail;
		if (new_inode->i_nlink == 0) {
			ret = btrfs_orphan_add(trans, new_dentry->d_inode);
			if (ret)
				goto out_fail;
		}

	}
	ret = btrfs_set_inode_index(new_dir, &index);
	if (ret)
		goto out_fail;

	ret = btrfs_add_link(trans, new_dentry->d_parent->d_inode,
			     old_inode, new_dentry->d_name.name,
			     new_dentry->d_name.len, 1, index);
	if (ret)
		goto out_fail;

	btrfs_log_new_name(trans, old_inode, old_dir,
				       new_dentry->d_parent);
out_fail:

	/* this btrfs_end_log_trans just allows the current
	 * log-sub transaction to complete
	 */
	btrfs_end_log_trans(root);
	btrfs_end_transaction_throttle(trans, root);
out_unlock:
	return ret;
}

/*
 * some fairly slow code that needs optimization. This walks the list
 * of all the inodes with pending delalloc and forces them to disk.
 */
int btrfs_start_delalloc_inodes(struct btrfs_root *root)
{
	struct list_head *head = &root->fs_info->delalloc_inodes;
	struct btrfs_inode *binode;
	struct inode *inode;

	if (root->fs_info->sb->s_flags & MS_RDONLY)
		return -EROFS;

	spin_lock(&root->fs_info->delalloc_lock);
	while (!list_empty(head)) {
		binode = list_entry(head->next, struct btrfs_inode,
				    delalloc_inodes);
		inode = igrab(&binode->vfs_inode);
		if (!inode)
			list_del_init(&binode->delalloc_inodes);
		spin_unlock(&root->fs_info->delalloc_lock);
		if (inode) {
			filemap_flush(inode->i_mapping);
			iput(inode);
		}
		cond_resched();
		spin_lock(&root->fs_info->delalloc_lock);
	}
	spin_unlock(&root->fs_info->delalloc_lock);

	/* the filemap_flush will queue IO into the worker threads, but
	 * we have to make sure the IO is actually started and that
	 * ordered extents get created before we return
	 */
	atomic_inc(&root->fs_info->async_submit_draining);
	while (atomic_read(&root->fs_info->nr_async_submits) ||