| To: | Thiemo Seufer <ths@networkno.de> |
|---|---|
| Subject: | Re: [PATCH] NEC EMMA2RH support |
| From: | Ralf Baechle <ralf@linux-mips.org> |
| Date: | Fri, 19 May 2006 11:52:52 +0100 |
| Cc: | dmitry pervushin <dpervushin@ru.mvista.com>, linux-mips@linux-mips.org |
| In-reply-to: | <20060518111703.GA15601@networkno.de> |
| Original-recipient: | rfc822;linux-mips@linux-mips.org |
| References: | <1147946423.8223.4.camel@diimka-laptop> <20060518195404.663eba86.yoichi_yuasa@tripeaks.co.jp> <1147950509.8223.10.camel@diimka-laptop> <20060518111703.GA15601@networkno.de> |
| Sender: | linux-mips-bounce@linux-mips.org |
| User-agent: | Mutt/1.4.2.1i |
On Thu, May 18, 2006 at 12:17:04PM +0100, Thiemo Seufer wrote: > > May be, I have > > misunderstood the modern ways in linux kernel development, but I am > > pretty sure that assembler interrupt handler will be faster than C > > code. > > Only marginally, it doesn't outweigh the maintenance trouble. Actually the average interrupt handler was sufficiently badly scheduled such that the C written ones were usually better. Not only, gcc knows alot of CPU specifics about scheduling, so the average interrupt should now suffer from many less taken branches. Ralf |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | I2C troubles with Au1550, Clem Taylor |
|---|---|
| Next by Date: | Re: Can't debug core files with GDB, Daniel Jacobowitz |
| Previous by Thread: | Re: [PATCH] NEC EMMA2RH support, Thiemo Seufer |
| Next by Thread: | Re: [PATCH] NEC EMMA2RH support, Ralf Baechle |
| Indexes: | [Date] [Thread] [Top] [All Lists] |