From mboxrd@z Thu Jan 1 00:00:00 1970 From: Timur Tabi Date: Fri, 04 May 2007 11:32:34 -0500 Subject: [U-Boot-Users] [PATCH v3] Fix initrd length miscalculation in bootm command when using a dtu In-Reply-To: <20070504162710.691AE353AF1@atlas.denx.de> References: <20070504162710.691AE353AF1@atlas.denx.de> Message-ID: <463B6022.7000908@freescale.com> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de Wolfgang Denk wrote: > In message <463B3463.1030702@freescale.com> you wrote: >> If you apply my patch, then the comment becomes correct. My goal was to >> lock the variables 'len' and 'data' into one purpose. The reason the > > I think you actually missed a few places where len was also used. Really? I doubt it. Every other place that len is used, it's the initrd length. Can you show me where that's not the case? >> bug existed is because the other developer didn't realize this. He used >> 'len' thinking it was available. In a sense, I was trying to implement > > This type of thinking was plain wrong. You must never use any > variable without understanding what it's being used for. You're absolutely right, which is why *I* don't make that mistake. But some one else did, and someone else might again. do_bootm_linux() is hideous code, and it just gets worse and worse. Frankly, anyone trying to modify that code needs all the help he can get! > As I explained before, you need a normal developer's repo, not a > guardian one. And I repeat: I offer you such a repo, it's available > right now, I just need the IP address from where you will connect. It would be the same IP address that Kim, Andy, and Jon use. As soon as I get hold of one of them, I'll ask, but technically you should know the answer already. -- Timur Tabi Linux Kernel Developer @ Freescale