From mboxrd@z Thu Jan 1 00:00:00 1970 From: Scott Wood Date: Wed, 20 Feb 2013 12:12:23 -0600 Subject: [U-Boot] [PATCH v4] imls: Add support to list images in NAND device In-Reply-To: <512444F3.6010109@st.com> (from vipin.kumar@st.com on Tue Feb 19 21:37:23 2013) Message-ID: <1361383943.31212.3@snotra> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de On 02/19/2013 09:37:23 PM, Vipin Kumar wrote: > On 2/20/2013 6:45 AM, Scott Wood wrote: >> On 12/17/2012 02:32:48 AM, Vipin Kumar wrote: >>> +#if defined(CONFIG_CMD_IMLS_NAND) >>> +static int nand_imls_legacyimage(nand_info_t *nand, int nand_dev, >>> loff_t off, >>> + size_t len) >>> +{ >>> + void *imgdata; >>> + int ret; >>> + >>> + imgdata = malloc(len); >>> + if (!imgdata) { >>> + printf("May be a Legacy Image at NAND device %d offset >>> %08llX:\n", >>> + nand_dev, off); >>> + printf(" Low memory(cannot allocate memory for >>> image)\n"); >>> + return -ENOMEM; >>> + } >> >> So, it looks like any reasonably-sized kernel image is going to hit >> this out-of-memory condition given how unreasonably small U-Boot >> malloc >> heaps generally are. >> > > Yes, that's right. And it is a pity I think I'll apply this anyway, as it is still helpful, but a future enhancement could be to do a streaming read of the image to verify the CRC. >> As a side note, running this showed U-Boot itself as "may be a legacy >> image"... it seems that for some reason all the ppc arches start with >> the uImage header, but otherwise aren't valid uImages. > > This "may be a legacy image" print is when we have successfully > copied and validated the 64 byte mkimage header and are not able to > allocate the memory from heap for the image data. What it means is > that although the whole image might be a good "Legacy image" but we > can not guarantee because we have not yet validated the dcrc. > > Thats why this print "May be a legacy image" Sure, that comment was meant as "hmm, powerpc is doing something weird" rather than anything wrong with this patch. -Scott