This is the mail archive of the
mailing list for the glibc project.
Re: [PATCH 4/6] vfprintf: Move jump table definition and the macros out of function
- From: "Carlos O'Donell" <carlos at redhat dot com>
- To: Florian Weimer <fweimer at redhat dot com>, libc-alpha at sourceware dot org
- Date: Fri, 06 Mar 2015 10:41:22 -0500
- Subject: Re: [PATCH 4/6] vfprintf: Move jump table definition and the macros out of function
- Authentication-results: sourceware.org; auth=none
- References: <cover dot 1425246936 dot git dot fweimer at redhat dot com> <f8ad26637a01b8173077d6dd38dd13cc31bbf22d dot 1425246936 dot git dot fweimer at redhat dot com> <54F8B0C1 dot 8010205 at redhat dot com> <54F98300 dot 2040100 at redhat dot com>
On 03/06/2015 05:35 AM, Florian Weimer wrote:
> On 03/05/2015 08:38 PM, Carlos O'Donell wrote:
>> On 03/01/2015 04:00 PM, Florian Weimer wrote:
>>> The second jump table will be moved to a separate function
>>> in the next commit.
>>> stdio-common/vfprintf.c | 108 ++++++++++++++++++++++++------------------------
>>> 1 file changed, 54 insertions(+), 54 deletions(-)
>> How does this impact the generated code and performance of these routines?
> The generated code is identical.
> Or is your question about next change (the printf_positional function)?
The question was for this specific change, but it does apply equally
across the board. I see your other post about the fact that it actually
improves performance and I'll reply there.