Any updates about the fallocate() issue I've described (I sent the related patch) below reported ?
Thanks in advance
From: Francesco RUNDO [mailto:francesco.rundo@...]
Sent: Monday, November 22, 2010 3:41 PM
Subject: [LTP] Patch to cover fallocate issue
I'm using latest LTP-full-20101031 cross-built for SH based platforms.
I'd like to submit a fix related to "fallocate" testcase (fallocate01,fallocate02,fallocate03).
Currently, the testcase checks the fallocate syscall also in case of "such" (and not "all") architectures with wordsize=32: ppc64, ppc, x86_64. Clearly, the testcase manages the parameters -for the fallocate syscall- according to the endianess supported by the above archs.
The so developed testcase will fail for all the architectures -with wordsize=32- and/or different endianess, with respect to the ones (board and/or endianess) above described and currently supported in the tests (for instance, the test failed in case of execution on SH based platforms!).
Please, find in attach a patch which aim is to extend the fallocate testcase(all subtests: fallocate01,fallocate02,fallocate03) to all the architectures with wordsize=32 both big and little endian.
From: Garrett Cooper <yanegomi@gm...> - 2010-12-08 09:00:59
On Tue, Nov 30, 2010 at 5:10 AM, Francesco RUNDO <francesco.rundo@...> wrote:
> Any updates about the fallocate() issue I’ve described (I sent the related
> patch) below reported ?
Pretty sure this breaks 32-bit archs.