Re: read vs fread
From: David Schwartz (davids_at_webmaster.com)
Date: 02/25/05
- Next message: Jan Kandziora: "Re: shared libs"
- Previous message: Claudio Takahasi: "Bluetooth and Wi-Fi for OMAP1510"
- In reply to: Grant Edwards: "Re: read vs fread"
- Next in thread: Patrice Kadionik: "Re: read vs fread"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Date: Fri, 25 Feb 2005 12:03:22 -0800
"Grant Edwards" <grante@visi.com> wrote in message
news:421f4794$0$46582$a1866201@visi.com...
> On 2005-02-25, Jan Kandziora <jjj@gmx.de> wrote:
>> Patrice Kadionik schrieb:
>
>>> - are syscalls.
>>
>> No, they are library functions of libc. Syscalls are special calls, which
>> use the Kernelspace/Userspace gate. They are implemented e.g. by software
>> interrupts in Linux on the x86 platform. These are used *inside* the
>> read()
>> etc. functions.
>
> That's technically true, but those library calls map directly
> into syscalls (same parameters, same return values) and are
> documented in a section 2 of the manual (generally referred to
> as the system call section) instead of section 3 (generally
> referred to as library section).
Except that can have huge semantic differences. On LinuxThreads, for
example, the system calls are not cancellation points, but the library
functions are.
DS
- Next message: Jan Kandziora: "Re: shared libs"
- Previous message: Claudio Takahasi: "Bluetooth and Wi-Fi for OMAP1510"
- In reply to: Grant Edwards: "Re: read vs fread"
- Next in thread: Patrice Kadionik: "Re: read vs fread"
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Relevant Pages
|