From: Junio C Hamano <gitster@pobox.com>
Date: Mon, 23 Apr 2012 21:13:51 GMT
Subject: Re: [PATCH 1/3] git-svn: use POSIX::sigprocmask to block signals
Message-ID: <xmqqd36yi228.fsf@junio.mtv.corp.google.com>
In-Reply-To: <CANiYKX68AJyE3+3BT7QSjziGS=HGKuk95rxS_04c0sGz9pywXw@mail.gmail.com>

Roman Kagan <rkagan@mail.ru> writes:

> [B]locking some signals is obviously not a good enough
> protection measure: the program may die on SIGKILL or the power may go
> off.

Ahh, that is what you meant.  I agree that it is of dubious value to
only try to catch signals, even though it may be better than nothing.

It should be sufficient to replace the "[...]" part with the above,
and conclude it with "But it will be a separate topic that this patch
does not address." or something.

Thanks.