MDEV-39123: Replace sprintf with snprintf in the InnoDB binlog code

[gkodinov]

sprintf is deprecated (and triggers a deprecation warning) on some platforms, saying it needs to be replaced with snprintf.
This is a good idea, since sprintf doesn't explicitly check the size of the buffer it is printing to.
Fixed by adding an extra size argument to the helper function and making sure this extra argument is passed down from the caller to snprintf.

[vaintroub]

sizeof_name_buf, really? maybe there is something more natural, and short, "buf_size"?

[gkodinov]

used name_buf_size.

[sanja-byelkin]

see bb-10.11-mac and Marko aproaches in the correspondent MDEV

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.

---

Georgi (Joro) Kodinov seems not to be a GitHub user. You need a GitHub account to be able to sign the CLA. If you have already a GitHub account, please add the email address used for this commit to your account.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[dr-m]

This function is a simple wrapper of snprintf(). What happens if more than BINLOG_NAME_MAX_LEN characters of input is available? Would the filename be terminated by \0, or could the subsequent sql_print_information() call exceed the bounds of the filename buffer. I tested this, because the documentation of snprintf(3) on my system is a little unclear:

#include <stdio.h>
int main()
{
  char buf[2];
  snprintf(buf, sizeof buf, "foo");
  return printf("hello %s\n", buf);
}

On my system, this would display hello f, so this appears to be fine.

There used to be an issue on Microsoft Windows here, but as noted in https://stackoverflow.com/questions/7706936/is-snprintf-always-null-terminating it was corrected some 10 years ago by implementing the standard snprintf() function.

[gkodinov]

https://linux.die.net/man/3/snprintf says:

The functions snprintf() and vsnprintf() write at most size bytes (including the terminating null byte ('\0')) to str.

[dr-m]

You could shoot at the static_assert message as well and correct the shot to short. Given that this branch includes cacaaeb, we could use the unary form of static_assert and reduce the source code line count by 1.

[gkodinov]

Please elaborate.

[dr-m]

There is a typo "too shot" in the static_assert message. You could just write

  static_assert(BINLOG_NAME_MAX_LEN <= FN_REFLEN);

because this branch uses a C++ standard version that supports the unary variant of static_assert.

[gkodinov]

a bit outside of the scope of the fix, but OK. fixing.

[dr-m]

The parentheses after sizeof are only needed when the argument is a name of a type. Here it is an expression, and therefore the parentheses are redundant.

[gkodinov]

I will remove the parentheses. But, I was wondering (since this is not wrong as is), is there any coding style description that was violated? Or is this just a personal preference?