"Dr. David Alan Gilbert" <[email protected]> writes:
> * Fabiano Rosas ([email protected]) wrote:
>> The following patches will touch HMP readline completion for migration
>> parameters, add a test case.
>>
>> The test puts the monitor on a chardev via socket and bypasses qtest
>> facilities because it needs to emit raw codes to readline. It
>> therefore requires a couple of new helpers to read/write to the
>> monitor socket.
>>
>> Usage:
>> QTEST_QEMU_BINARY=./qemu-system-x86_64 \
>> ./tests/qtest/migration-test --full -p /x86_64/migration/hmp/completion
>>
>> Signed-off-by: Fabiano Rosas <[email protected]>
>> ---
>
>> +static void hmp_sock_read(int fd, char *buf, size_t buf_sz)
>> +{
>> + char *p = buf;
>> + size_t sz = buf_sz - 1;
>> +
>> + assert(fd >= 0);
>> + memset(buf, 0, buf_sz);
>> +
>> + while (sz > 0) {
>> + ssize_t r = read(fd, p, sz);
>> + char *prompt;
>> +
>> + if (!r) {
>> + break;
>> + } else if (r < 0) {
>> + if (errno == EINTR) {
>> + continue;
>> + }
>> + g_assert_not_reached();
>> + }
>> +
>> + p += r;
>> + sz -= r;
>> +
>> + prompt = strstr(buf, "(qemu) ");
>> + if (prompt) {
>> + *prompt = '\0';
>
> I'm a bit confused by what you want the output to look like after this,
> are you assuming this is the end of the string you're reading so truncation
> is OK?
>
Yes, it's how I'm detecting the end of input. Here's a printf("%s|\n", buf)
before the truncation:
1) Read banner:
QEMU 11.1.50 monitor - type 'help' for more information
(qemu) |
2) Write completion input 'migra\t'. Then read back:
migra
migrate migrate_cancel migrate_continue
migrate_incoming migrate_pause migrate_recover
migrate_set_capability migrate_set_parameter migrate_start_postcopy
(qemu) |
3) Write completion input 'info migra\t'. Then read back:
info migra
migrate migrate_capabilities migrate_parameters
(qemu) |
I think this is correct? Let me know if I missed something.
>> + break;
>> + }
>> + }
>> +}
>> +
>> +static int comp(const void *a, const void *b)
>> +{
>> + return strcmp(*(const char **) a, *(const char **) b);
>> +}
>
> Can you use qemu_pstrcmp0 there? (Maybe with g_ptr_array and g_ptr_array_sort
> is
> what adds to some of the other questions below?)
>
Good idea, specially the g_ptr_array. I'll update the code.
