summarylogtreecommitdiffstats
path: root/vmmon.patch
blob: 16dc1755c71ebd675f17454a9c419ad7093441e0 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
--- a/vmmon/Makefile
+++ b/vmmon/Makefile
@@ -43,7 +43,11 @@ INCLUDE      += -I$(SRCROOT)/shared
 endif
 
 
+ifdef KVERSION
+VM_UNAME = $(KVERSION)
+else
 VM_UNAME = $(shell uname -r)
+endif
 
 # Header directory for the running kernel
 ifdef LINUXINCLUDE
@@ -100,6 +104,13 @@ auto-build: $(DRIVER_KO)
 $(DRIVER): $(DRIVER_KO)
 	if [ $< -nt $@ ] || [ ! -e $@ ] ; then cp -f $< $@; fi
 
+# Use SUBDIRS on 2.x, 3.x, 4.x.  Use M on newer kernels.
+ifeq ($(filter-out 2 3 4,$(firstword $(subst ., ,$(VM_UNAME)))),)
+DIRVAR := SUBDIRS
+else
+DIRVAR := M
+endif
+
 # Pass gcc version down the chain, so we can detect if kernel attempts to use unapproved compiler
 VM_CCVER := $(VMCCVER)
 export VM_CCVER
@@ -117,7 +128,7 @@ prebuild:: ;
 postbuild:: ;
 
 $(DRIVER_KO): prebuild
-	$(MAKE) -C $(BUILD_DIR) SUBDIRS=$$PWD SRCROOT=$$PWD/$(SRCROOT) \
+	$(MAKE) -C $(BUILD_DIR) $(DIRVAR)=$$PWD SRCROOT=$$PWD/$(SRCROOT) \
 	  MODULEBUILDDIR=$(MODULEBUILDDIR) modules
 	$(MAKE) -C $$PWD SRCROOT=$$PWD/$(SRCROOT) \
 	  MODULEBUILDDIR=$(MODULEBUILDDIR) postbuild
--- a/vmmon/linux/driver.c
+++ b/vmmon/linux/driver.c
@@ -80,6 +80,16 @@
 
 struct VMXLinuxState linuxState;
 
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(5, 0, 0)
+static inline void do_gettimeofday(struct timeval *tv)
+{
+   struct timespec64 now;
+
+   ktime_get_real_ts64(&now);
+   tv->tv_sec = now.tv_sec;
+   tv->tv_usec = now.tv_nsec / 1000;
+}
+#endif
 
 /*
  *----------------------------------------------------------------------
@@ -104,7 +114,9 @@ long LinuxDriver_Ioctl(struct file *filp, u_int iocmd,
 
 static int LinuxDriver_Close(struct inode *inode, struct file *filp);
 static unsigned int LinuxDriverPoll(struct file *file, poll_table *wait);
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 11, 0)
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 17, 0)
+static vm_fault_t LinuxDriverFault(struct vm_fault *fault);
+#elif LINUX_VERSION_CODE >= KERNEL_VERSION(4, 11, 0)
 static int LinuxDriverFault(struct vm_fault *fault);
 #elif defined(VMW_NOPAGE_2624)
 static int LinuxDriverFault(struct vm_area_struct *vma, struct vm_fault *fault);
@@ -115,7 +115,7 @@
 #endif
 static int LinuxDriverMmap(struct file *filp, struct vm_area_struct *vma);
 
-static void LinuxDriverPollTimeout(unsigned long clientData);
+static void LinuxDriverPollTimeout(struct timer_list *clientData);
 static unsigned int LinuxDriverEstimateTSCkHz(void);
 
 static struct vm_operations_struct vmuser_mops = {
@@ -227,7 +228,7 @@ LinuxDriverEstimateTSCkHz(void)
  *----------------------------------------------------------------------
  */
 static void
-LinuxDriverEstimateTSCkHzDeferred(unsigned long data)
+LinuxDriverEstimateTSCkHzDeferred(struct timer_list *data)
 {
    LinuxDriverEstimateTSCkHz();
 }
@@ -265,9 +266,7 @@ LinuxDriverInitTSCkHz(void)
    }
 
    Vmx86_ReadTSCAndUptime(&tsckHzStartTime);
-   tscTimer.function = LinuxDriverEstimateTSCkHzDeferred;
    tscTimer.expires  = jiffies + 4 * HZ;
-   tscTimer.data     = 0;
    add_timer(&tscTimer);
 }
 
@@ -309,9 +308,13 @@ init_module(void)
     */
 
    init_waitqueue_head(&linuxState.pollQueue);
+#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 15, 0) && !defined(timer_setup)
    init_timer(&linuxState.pollTimer);
    linuxState.pollTimer.data = 0;
-   linuxState.pollTimer.function = LinuxDriverPollTimeout;
+   linuxState.pollTimer.function = (void *)LinuxDriverPollTimeout;
+#else
+   timer_setup(&linuxState.pollTimer, LinuxDriverPollTimeout, 0);
+#endif
 
    linuxState.fastClockThread = NULL;
    linuxState.fastClockFile = NULL;
@@ -360,7 +357,13 @@ init_module(void)
        linuxState.deviceName, linuxState.major, linuxState.minor);
 
    HostIF_InitUptime();
+#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 15, 0) && !defined(timer_setup)
    init_timer(&tscTimer);
+   tscTimer.function = (void *)LinuxDriverEstimateTSCkHzDeferred;
+   tscTimer.data = (unsigned long)&tscTimer;
+#else
+   timer_setup(&tscTimer, LinuxDriverEstimateTSCkHzDeferred, 0);
+#endif
    LinuxDriverInitTSCkHz();
    Vmx86_InitIDList();
 
@@ -858,7 +858,7 @@
  */
 
 static void
-LinuxDriverPollTimeout(unsigned long clientData)  // IN:
+LinuxDriverPollTimeout(struct timer_list *clientData)  // IN:
 {
    LinuxDriverWakeUp(FALSE);
 }
@@ -884,7 +904,12 @@ LinuxDriver_Close(struct inode *inode, // IN
  */
 
 #if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 11, 0)
-static int
+static
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 17, 0)
+vm_fault_t
+#else
+int
+#endif
 LinuxDriverFault(struct vm_fault *fault)     //IN/OUT
 #elif defined(VMW_NOPAGE_2624)
 static int LinuxDriverFault(struct vm_area_struct *vma, //IN
--- a/vmmon/linux/hostif.c
+++ b/vmmon/linux/hostif.c
@@ -1164,16 +1172,7 @@
 {
    int retval;
 
-   down_read(&current->mm->mmap_sem);
-#if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 9, 0)
-   retval = get_user_pages((unsigned long)uvAddr, numPages, 0, ppages, NULL);
-#elif LINUX_VERSION_CODE >= KERNEL_VERSION(4, 6, 0)
-   retval = get_user_pages((unsigned long)uvAddr, numPages, 0, 0, ppages, NULL);
-#else
-   retval = get_user_pages(current, current->mm, (unsigned long)uvAddr,
-                           numPages, 0, 0, ppages, NULL);
-#endif
-   up_read(&current->mm->mmap_sem);
+   retval = get_user_pages_fast((unsigned long)uvAddr, numPages, 0, ppages);
 
    return retval != numPages;
 }
@@ -1574,9 +1565,13 @@
     * since at least 2.6.0.
     */
 
+#if LINUX_VERSION_CODE < KERNEL_VERSION(5, 0, 0)
    extern unsigned long totalram_pages;
 
    unsigned int totalPhysicalPages = totalram_pages;
+#else
+   unsigned int totalPhysicalPages = totalram_pages();
+#endif
 
 #if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 28)
    return MemDefaults_CalcMaxLockedPages(totalPhysicalPages);
@@ -1594,17 +1593,34 @@ HostIF_EstimateLockedPageLimit(const VMDriver* vm,                // IN
    unsigned int reservedPages = MEMDEFAULTS_MIN_HOST_PAGES;
    unsigned int hugePages = (vm == NULL) ? 0 :
       BYTES_2_PAGES(vm->memInfo.hugePageBytes);
-   unsigned int lockedPages = global_page_state(NR_PAGETABLE) +
-                              global_page_state(NR_SLAB_UNRECLAIMABLE) +
-                              global_page_state(NR_UNEVICTABLE) +
-                              hugePages + reservedPages;
-   unsigned int anonPages =
+   unsigned int lockedPages = hugePages + reservedPages;
+   unsigned int anonPages;
+   unsigned int swapPages = BYTES_2_PAGES(linuxState.swapSize);
+
+   /* global_page_state is global_zone_page_state in 4.14. */
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 14, 0)
+   lockedPages += global_zone_page_state(NR_PAGETABLE);
+#else
+   lockedPages += global_page_state(NR_PAGETABLE);
+#endif
+   /* NR_SLAB_* moved from zone to node in 4.13. */
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 13, 0)
+   lockedPages += global_node_page_state(NR_SLAB_UNRECLAIMABLE);
+#else
+   lockedPages += global_page_state(NR_SLAB_UNRECLAIMABLE);
+#endif
+   /* NR_UNEVICTABLE moved from global to node in 4.8. */
 #if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 8, 0)
-      global_page_state(NR_ANON_MAPPED);
+   lockedPages += global_node_page_state(NR_UNEVICTABLE);
 #else
-      global_page_state(NR_ANON_PAGES);
+   lockedPages += global_page_state(NR_UNEVICTABLE);
+#endif
+   /* NR_ANON_MAPPED moved & changed name in 4.8. */
+#if LINUX_VERSION_CODE >= KERNEL_VERSION(4, 8, 0)
+   anonPages = global_node_page_state(NR_ANON_MAPPED);
+#else
+   anonPages = global_page_state(NR_ANON_PAGES);
 #endif
-   unsigned int swapPages = BYTES_2_PAGES(linuxState.swapSize);
 
    if (anonPages > swapPages) {
       lockedPages += anonPages - swapPages; 
@@ -1664,6 +1676,49 @@
 /*
  *----------------------------------------------------------------------
  *
+ * HostIFGetTime --
+ *
+ *      Reads the current time in UPTIME_FREQ units.
+ *
+ * Results:
+ *      The uptime, in units of UPTIME_FREQ.
+ *
+ * Side effects:
+ *      None.
+ *
+ *----------------------------------------------------------------------
+ */
+
+static uint64
+HostIFGetTime(void)
+{
+#if LINUX_VERSION_CODE < KERNEL_VERSION(5, 0, 0)
+   struct timeval tv;
+
+   do_gettimeofday(&tv);
+   return tv.tv_usec * (UPTIME_FREQ / 1000000) + tv.tv_sec * UPTIME_FREQ;
+#else
+   struct timespec64 now;
+
+   /*
+    * Use raw time used by Posix timers.  This time is not affected by
+    * NTP adjustments, so it may drift from real time and monotonic time,
+    * but it will stay in sync with other timers.
+    */
+   ktime_get_raw_ts64(&now);
+   /*
+    * UPTIME_FREQ resolution is lower than tv_nsec,
+    * so we have to do division...
+    */
+   ASSERT_ON_COMPILE(1000000000 % UPTIME_FREQ == 0);
+   return now.tv_nsec / (1000000000 / UPTIME_FREQ) + now.tv_sec * UPTIME_FREQ;
+#endif
+}
+
+
+/*
+ *----------------------------------------------------------------------
+ *
  * HostIFReadUptimeWork --
  *
  *      Reads the current uptime.  The uptime is based on getimeofday,
@@ -1692,16 +1747,12 @@
 static uint64
 HostIFReadUptimeWork(unsigned long *j)  // OUT: current jiffies 
 {
-   struct timeval tv;
    uint64 monotime, uptime, upBase, monoBase;
    int64 diff;
    uint32 version;
    unsigned long jifs, jifBase;
    unsigned int attempts = 0;
 
-   /* Assert that HostIF_InitUptime has been called. */
-   ASSERT(uptimeState.timer.function);
-
  retry:
    do {
       version  = VersionedAtomic_BeginTryRead(&uptimeState.version);
@@ -1710,13 +1761,12 @@
       monoBase = uptimeState.monotimeBase;
    } while (!VersionedAtomic_EndTryRead(&uptimeState.version, version));
 
-   do_gettimeofday(&tv);
+   uptime = HostIFGetTime();
    upBase = Atomic_Read64(&uptimeState.uptimeBase);
    
    monotime = (uint64)(jifs - jifBase) * (UPTIME_FREQ / HZ);
    monotime += monoBase;
 
-   uptime = tv.tv_usec * (UPTIME_FREQ / 1000000) + tv.tv_sec * UPTIME_FREQ;
    uptime += upBase;
    
    /* 
@@ -1767,7 +1804,7 @@ HostIFReadUptimeWork(unsigned long *j)  // OUT: current jiffies
  */
 
 static void
-HostIFUptimeResyncMono(unsigned long data)  // IN: ignored
+HostIFUptimeResyncMono(struct timer_list *timer)  // IN: ignored
 {
    unsigned long jifs;
    uintptr_t flags;
@@ -1821,16 +1871,19 @@
 void
 HostIF_InitUptime(void)
 {
-   struct timeval tv;
+   uint64 tm;
 
    uptimeState.jiffiesBase = jiffies;
-   do_gettimeofday(&tv);
-   Atomic_Write64(&uptimeState.uptimeBase, 
-                  -(tv.tv_usec * (UPTIME_FREQ / 1000000) + 
-                    tv.tv_sec * UPTIME_FREQ));
+   tm = HostIFGetTime();
+   Atomic_Write64(&uptimeState.uptimeBase, -tm);
 
+#if LINUX_VERSION_CODE < KERNEL_VERSION(4, 15, 0) && !defined(timer_setup)
    init_timer(&uptimeState.timer);
-   uptimeState.timer.function = HostIFUptimeResyncMono;
+   uptimeState.timer.function = (void *)HostIFUptimeResyncMono;
+   uptimeState.timer.data = (unsigned long)&uptimeState.timer;
+#else
+   timer_setup(&uptimeState.timer, HostIFUptimeResyncMono, 0);
+#endif
    mod_timer(&uptimeState.timer, jiffies + HZ);
 }
 
@@ -2251,7 +2304,7 @@ isVAReadable(VA r)  // IN:
    int ret;
    
    old_fs = get_fs();
-   set_fs(get_ds());
+   set_fs(KERNEL_DS);
    r = APICR_TO_ADDR(r, APICR_VERSION);
    ret = HostIF_CopyFromUser(&dummy, (void*)r, sizeof(dummy));
    set_fs(old_fs);
@@ -2452,7 +2505,7 @@ HostIF_SemaphoreWait(VMDriver *vm,   // IN:
    }
 
    old_fs = get_fs();
-   set_fs(get_ds());
+   set_fs(KERNEL_DS);
 
    {
       struct poll_wqueues table;
@@ -2581,7 +2634,7 @@ HostIF_SemaphoreSignal(uint64 *args)  // IN:
    }
 
    old_fs = get_fs();
-   set_fs(get_ds());
+   set_fs(KERNEL_DS);
 
    /*
     * Always write sizeof(uint64) bytes. This works fine for eventfd and
@@ -3524,7 +3577,6 @@ HostIF_SetFastClockRate(unsigned int rate) // IN: Frequency in Hz.
       }
    } else {
       if (linuxState.fastClockThread) {
-         force_sig(SIGKILL, linuxState.fastClockThread);
          kthread_stop(linuxState.fastClockThread);
 	 close_rtc(linuxState.fastClockFile, current->files);
 
@@ -3572,7 +3625,12 @@
 
    ASSERT(handle);
 
-   if (!access_ok(VERIFY_WRITE, p, size)) {
+#if LINUX_VERSION_CODE < KERNEL_VERSION(5, 0, 0)
+   if (!access_ok(VERIFY_WRITE, p, size))
+#else
+   if (!access_ok(p, size))
+#endif
+   {
       printk(KERN_ERR "%s: Couldn't verify write to uva 0x%p with size %"
              FMTSZ"u\n", __func__, p, size);