[Top] [All Lists]

Users pace crash on Netlogic XLP II processors

To: "" <>
Subject: Users pace crash on Netlogic XLP II processors
From: Jacky <>
Date: Tue, 26 Nov 2013 08:44:56 +0800
Cc: "" <>
Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed;; s=20120113; h=subject:mime-version:from:content-type:message-id:date:cc :content-transfer-encoding:to; bh=BUuGooKyVBY+FuTNjECa8PX8f9FHe5ca0+Yh4yfgeRc=; b=DcM+8MTiPJI9EL1KevkCCY8vGU1W4AAr2/4x8E873inGG/k07Q6I8dWDTUni2fdzgA TM+DP/Cm4SAW09KtKcEZvdcia+kBV7zav0DaHUUdxnkzWJcaoEcozadiU8SbgJmNZbsk 1ep/fbFpTagjMzQN8oZ9cHHBbxnKyF8xRrpsFwPpDWfX+W4ohkYLE/JUxHvpNglMz039 VWXkJ9g+cFbKuGQSQbI9+RA/5WLXhJp+8cj9bYckzq/b99CNFCzgweLHVjAvrhp72PGN /S7cj1BcdCLNznWnGUOvHxQ0PkIp7LOzmsGF/Km9rUy3J7HuI21kWc/xu0u1CkmFyx8Z 932g==
List-archive: <>
List-help: <>
List-id: linux-mips <>
List-owner: <>
List-post: <>
List-software: Ecartis version 1.0.0
List-subscribe: <>
List-unsubscribe: <>
Original-recipient: rfc822;
Recently I experienced user space crash on kernel 2.6.32 after executables and 
memory foot print grow over particular size. As I found the crash reason is 
inconsistent I-cache which the code is correct in memory (or L2) but definitely 
wrong in L1 I-cache.

Should I-cache invalidate happen during page reclaim? And Which cache flush 
hook responsible for it? The work around is ugly need invalidate I cache on 
every page fill.
<Prev in Thread] Current Thread [Next in Thread>
  • Users pace crash on Netlogic XLP II processors, Jacky <=