[Libguestfs] [libnbd PATCH v3 03/29] socket activation: rename sa_(tmpdir|sockpath) to sact_(tmpdir|sockpath)

[Laszlo Ersek]

On 2/15/23 21:00, Eric Blake wrote:
> On Wed, Feb 15, 2023 at 03:11:32PM +0100, Laszlo Ersek wrote:
>> <signal.h> in POSIX reserves the "sa_" prefix:
>>
>> https://pubs.opengroup.org/onlinepubs/9699919799/functions/V2_chap02.html#tag_15_02_02
>>
>> Let's use "sact_" instead.
>>
>> Signed-off-by: Laszlo Ersek <lersek at redhat.com>
>> ---
>>  lib/internal.h                               |  4 ++--
>>  generator/states-connect-socket-activation.c | 16 ++++++++--------
>>  lib/handle.c                                 | 12 ++++++------
>>  3 files changed, 16 insertions(+), 16 deletions(-)
> 
> Reviewed-by: Eric Blake <eblake at redhat.com>
> 
> Stepping on reserved names is something that you can often get away
> with (if you don't conflict now, how likely is it that libc will
> actually introduce a future symbol that will conflict later?), so this
> is change is more on the theoretical side.  But I have no objections,
> and as you proved, it's a very easy mechanical fix once we even bother
> to identify that reserved name clashes could be a potential porting
> issue.  I've probably spent more time writing this email than you did
> in creating the patch proper (if you exclude the subsequent time
> testing that it builds and passes 'make check-valgrind'...)

I've actually reviewed (eyeballed) this patch myself *multiple times*,
very carefully. Mechanical changes can be deceptive :) Implementing them
seems easy, but their impact can be surprising.

In fact, whenever I need to review such a patch for someone else, I tend
to redo it myself from zero, then compare the patches (or the resultant
trees).

> Reviewed-by: Eric Blake <eblake at redhat.com>
> 

Thanks!
Laszlo