From mboxrd@z Thu Jan 1 00:00:00 1970 MIME-Version: 1.0 In-Reply-To: <77525476571d85282989a8a1ebe187a6@felloff.net> References: <77525476571d85282989a8a1ebe187a6@felloff.net> Date: Wed, 1 Apr 2015 10:51:05 +0100 Message-ID: From: Charles Forsyth To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net> Content-Type: multipart/alternative; boundary=f46d043c7ccc5db2990512a6a771 Subject: Re: [9fans] mips linker hoisting stack freeing operation Topicbox-Message-UUID: 4c737336-ead9-11e9-9d60-3106f5b1d025 --f46d043c7ccc5db2990512a6a771 Content-Type: text/plain; charset=UTF-8 On 1 April 2015 at 02:34, wrote: > theres my attempt at preventing the linker from doing so: It's one of those cases where it's funny it hasn't been noticed before, since the effects won't be confined to libdraw. Anyway, that change seems ok, although it's really a particular type of conflict (between set R29 and use R6 which depends on R29), but expressing that is tricky with the limited value tracking there, and the ADD R29 makes as good a use of the delay slot as any. --f46d043c7ccc5db2990512a6a771 Content-Type: text/html; charset=UTF-8 Content-Transfer-Encoding: quoted-printable

= On 1 April 2015 at 02:34, <cinap_lenrek@felloff.net>= wrote:
theres my attempt at preventing t= he linker from doing so:

It's one of those cases = where it's funny it hasn't been noticed before,
since the effects won't be confined to libdraw. Anyway, th= at change seems ok,
although it's r= eally a particular type of conflict (between set R29 and use R6 which depen= ds on R29),
but expressing that is tricky w= ith the limited value tracking there, and the ADD R29 makes
as good a use of the delay slot as any.
--f46d043c7ccc5db2990512a6a771--