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
|
package xgb
/*
Tests for XGB.
These tests only test the core X protocol at the moment. It isn't even
close to complete coverage (and probably never will be), but it does test
a number of different corners: requests with no replies, requests without
replies, checked (i.e., synchronous) errors, unchecked (i.e., asynchronous)
errors, and sequence number wrapping.
There are also a couple of benchmarks that show the difference between
correctly issuing lots of requests and gathering replies and
incorrectly doing the same. (This particular difference is one of the
claimed advantages of the XCB, and therefore XGB, family.
*/
import (
"fmt"
"log"
"math/rand"
"testing"
"time"
)
// The X connection used throughout testing.
var X *Conn
// init initializes the X connection, seeds the RNG and starts waiting
// for events.
func init() {
var err error
X, err = NewConn()
if err != nil {
log.Fatal(err)
}
rand.Seed(time.Now().UnixNano())
go grabEvents()
}
/******************************************************************************/
// Tests
/******************************************************************************/
// TestSynchronousError purposefully causes a BadLength error in an
// InternAtom request, and checks it synchronously.
func TestSynchronousError(t *testing.T) {
err := X.MapWindowChecked(0).Check() // resource id 0 is always invalid
if err == nil {
t.Fatalf("MapWindow: A MapWindow request that should return an " +
"error has returned a nil error.")
}
verifyMapWindowError(t, err)
}
// TestAsynchronousError does the same thing as TestSynchronousError, but
// grabs the error asynchronously instead.
func TestAsynchronousError(t *testing.T) {
X.MapWindow(0) // resource id 0 is always invalid
evOrErr := waitForEvent(t, 5)
if evOrErr.ev != nil {
t.Fatalf("After issuing an erroneous MapWindow request, we have "+
"received an event rather than an error: %s", evOrErr.ev)
}
verifyMapWindowError(t, evOrErr.err)
}
// TestCookieBuffer issues (2^16) + n requets *without* replies to guarantee
// that the sequence number wraps and that the cookie buffer will have to
// flush itself (since there are no replies coming in to flush it).
// And just like TestSequenceWrap, we issue another request with a reply
// at the end to make sure XGB is still working properly.
func TestCookieBuffer(t *testing.T) {
n := (1 << 16) + 10
for i := 0; i < n; i++ {
X.NoOperation()
}
TestProperty(t)
}
// TestSequenceWrap issues (2^16) + n requests w/ replies to guarantee that the
// sequence number (which is a 16 bit integer) will wrap. It then issues one
// final request to ensure things still work properly.
func TestSequenceWrap(t *testing.T) {
n := (1 << 16) + 10
for i := 0; i < n; i++ {
_, err := X.InternAtom(false, 5, "RANDO").Reply()
if err != nil {
t.Fatalf("InternAtom: %s", err)
}
}
TestProperty(t)
}
// TestProperty tests whether a random value can be set and read.
func TestProperty(t *testing.T) {
propName := randString(20) // whatevs
writeVal := randString(20)
readVal, err := changeAndGetProp(propName, writeVal)
if err != nil {
t.Error(err)
}
if readVal != writeVal {
t.Errorf("The value written, '%s', is not the same as the "+
"value read '%s'.", writeVal, readVal)
}
}
// TestWindowEvents creates a window, maps it, listens for configure notify
// events, issues a configure request, and checks for the appropriate
// configure notify event.
// This probably violates the notion of "test one thing and test it well,"
// but testing X stuff is unique since it involves so much state.
// Each request is checked to make sure there are no errors returned. If there
// is an error, the test is failed.
// You may see a window appear quickly and then disappear. Do not be alarmed :P
// It's possible that this test will yield a false negative because we cannot
// control our environment. That is, the window manager could override the
// placement set. However, we set override redirect on the window, so the
// window manager *shouldn't* touch our window if it is well-behaved.
func TestWindowEvents(t *testing.T) {
// The geometry to set the window.
gx, gy, gw, gh := 200, 400, 1000, 300
wid, err := X.NewId()
if err != nil {
t.Fatalf("NewId: %s", err)
}
screen := X.DefaultScreen() // alias
err = X.CreateWindowChecked(screen.RootDepth, wid, screen.Root,
0, 0, 500, 500, 0,
WindowClassInputOutput, screen.RootVisual,
CwBackPixel|CwOverrideRedirect, []uint32{0xffffffff, 1}).Check()
if err != nil {
t.Fatalf("CreateWindow: %s", err)
}
err = X.MapWindowChecked(wid).Check()
if err != nil {
t.Fatalf("MapWindow: %s", err)
}
// We don't listen in the CreateWindow request so that we don't get
// a MapNotify event.
err = X.ChangeWindowAttributesChecked(wid,
CwEventMask, []uint32{EventMaskStructureNotify}).Check()
if err != nil {
t.Fatalf("ChangeWindowAttributes: %s", err)
}
err = X.ConfigureWindowChecked(wid,
ConfigWindowX|ConfigWindowY|
ConfigWindowWidth|ConfigWindowHeight,
[]uint32{uint32(gx), uint32(gy), uint32(gw), uint32(gh)}).Check()
if err != nil {
t.Fatalf("ConfigureWindow: %s", err)
}
evOrErr := waitForEvent(t, 5)
switch event := evOrErr.ev.(type) {
case ConfigureNotifyEvent:
if event.X != int16(gx) {
t.Fatalf("x was set to %d but ConfigureNotify reports %d",
gx, event.X)
}
if event.Y != int16(gy) {
t.Fatalf("y was set to %d but ConfigureNotify reports %d",
gy, event.Y)
}
if event.Width != uint16(gw) {
t.Fatalf("width was set to %d but ConfigureNotify reports %d",
gw, event.Width)
}
if event.Height != uint16(gh) {
t.Fatalf("height was set to %d but ConfigureNotify reports %d",
gh, event.Height)
}
default:
t.Fatalf("Expected a ConfigureNotifyEvent but got %T instead.", event)
}
// Okay, clean up!
err = X.ChangeWindowAttributesChecked(wid,
CwEventMask, []uint32{0}).Check()
if err != nil {
t.Fatalf("ChangeWindowAttributes: %s", err)
}
err = X.DestroyWindowChecked(wid).Check()
if err != nil {
t.Fatalf("DestroyWindow: %s", err)
}
}
/******************************************************************************/
// Benchmarks
/******************************************************************************/
// BenchmarkInternAtomsGood shows how many requests with replies
// *should* be sent and gathered from the server. Namely, send as many
// requests as you can at once, then go back and gather up all the replies.
// More importantly, this approach can exploit parallelism better when
// GOMAXPROCS > 1.
// Run with `go test -run 'nomatch' -bench '.*' -cpu 1,2,6` if you have
// multiple cores to see the improvement that parallelism brings.
func BenchmarkInternAtomsGood(b *testing.B) {
b.StopTimer()
names := seqNames(b.N)
b.StartTimer()
cookies := make([]InternAtomCookie, b.N)
for i := 0; i < b.N; i++ {
cookies[i] = X.InternAtom(false, uint16(len(names[i])), names[i])
}
for _, cookie := range cookies {
cookie.Reply()
}
}
// BenchmarkInternAtomsBad shows how *not* to issue a lot of requests with
// replies. Namely, each subsequent request isn't issued *until* the last
// reply is made. This implies a round trip to the X server for every
// iteration.
func BenchmarkInternAtomsPoor(b *testing.B) {
b.StopTimer()
names := seqNames(b.N)
b.StartTimer()
for i := 0; i < b.N; i++ {
X.InternAtom(false, uint16(len(names[i])), names[i]).Reply()
}
}
/******************************************************************************/
// Helper functions
/******************************************************************************/
// changeAndGetProp sets property 'prop' with value 'val'.
// It then gets the value of that property and returns it.
// (It's used to check that the 'val' going in is the same 'val' going out.)
// It tests both requests with and without replies (GetProperty and
// ChangeProperty respectively.)
func changeAndGetProp(prop, val string) (string, error) {
propAtom, err := X.InternAtom(false, uint16(len(prop)), prop).Reply()
if err != nil {
return "", fmt.Errorf("InternAtom: %s", err)
}
typName := "UTF8_STRING"
typAtom, err := X.InternAtom(false, uint16(len(typName)), typName).Reply()
if err != nil {
return "", fmt.Errorf("InternAtom: %s", err)
}
err = X.ChangePropertyChecked(PropModeReplace, X.DefaultScreen().Root,
propAtom.Atom, typAtom.Atom, 8, uint32(len(val)), []byte(val)).Check()
if err != nil {
return "", fmt.Errorf("ChangeProperty: %s", err)
}
reply, err := X.GetProperty(false, X.DefaultScreen().Root, propAtom.Atom,
GetPropertyTypeAny, 0, (1<<32)-1).Reply()
if err != nil {
return "", fmt.Errorf("GetProperty: %s", err)
}
if reply.Format != 8 {
return "", fmt.Errorf("Property reply format is %d but it should be 8.",
reply.Format)
}
return string(reply.Value), nil
}
// verifyMapWindowError takes an error that is returned with an invalid
// MapWindow request with a window Id of 0 and makes sure the error is the
// right type and contains the correct values.
func verifyMapWindowError(t *testing.T, err error) {
switch e := err.(type) {
case WindowError:
if e.BadValue != 0 {
t.Fatalf("WindowError should report a bad value of 0 but "+
"it reports %d instead.", e.BadValue)
}
if e.MajorOpcode != 8 {
t.Fatalf("WindowError should report a major opcode of 8 "+
"(which is a MapWindow request), but it reports %d instead.",
e.MajorOpcode)
}
default:
t.Fatalf("Expected a WindowError but got %T instead.", e)
}
}
// randString generates a random string of length n.
func randString(n int) string {
byts := make([]byte, n)
for i := 0; i < n; i++ {
rando := rand.Intn(53)
switch {
case rando <= 25:
byts[i] = byte(65 + rando)
case rando <= 51:
byts[i] = byte(97 + rando - 26)
default:
byts[i] = ' '
}
}
return string(byts)
}
// seqNames creates a slice of NAME0, NAME1, ..., NAMEN.
func seqNames(n int) []string {
names := make([]string, n)
for i := range names {
names[i] = fmt.Sprintf("NAME%d", i)
}
return names
}
// evErr represents a value that is either an event or an error.
type evErr struct {
ev Event
err Error
}
// channel used to pass evErrs.
var evOrErrChan = make(chan evErr, 0)
// grabEvents is a goroutine that reads events off the wire.
// We used this instead of WaitForEvent directly in our tests so that
// we can timeout and fail a test.
func grabEvents() {
for {
ev, err := X.WaitForEvent()
evOrErrChan <- evErr{ev, err}
}
}
// waitForEvent asks the evOrErrChan channel for an event.
// If it doesn't get an event in 'n' seconds, the current test is failed.
func waitForEvent(t *testing.T, n int) evErr {
var evOrErr evErr
select {
case evOrErr = <-evOrErrChan:
case <-time.After(time.Second * 5):
t.Fatalf("After waiting 5 seconds for an event or an error, " +
"we have timed out.")
}
return evOrErr
}
|