Hi folks, I've been off for a pretty long time and I'm trying to resume a little bit. While testing HDT I found some serious bugs in HDT's CLI mode but also took the time to update the DMI detection tables. Please find my hw_update branch which is in sync with the current git tree. https://github.com/ErwanAliasr1/syslinux/commits/hw_update Erwan,
On Fri, Aug 14, 2015 at 4:41 AM, Erwan Velu via Syslinux <syslinux at zytor.com> wrote:> Hi folks, > > I've been off for a pretty long time and I'm trying to resume a little bit. > > While testing HDT I found some serious bugs in HDT's CLI mode but also took > the time to update the DMI detection tables. > > Please find my hw_update branch which is in sync with the current git tree. > https://github.com/ErwanAliasr1/syslinux/commits/hw_updatef249d92b doesn't feel right. After parsing, it should become an array of many pointers, one per argument. You do this allocation in parse_command_line() properly but I believe due to the indirection, it's not being applied correctly. Don't we need the address of (1) an array (2) of pointers (3) to blocks of chars? static void parse_command_line(char *line, char **command, char **module, int *argc, char **argv) should probably be static void parse_command_line(char *line, char **command, char **module, int *argc, char ***argv) with other appropriate changes while parse_command_line(line, &command, &module, &argc, argv); should be parse_command_line(line, &command, &module, &argc, &argv); Let me know. -- -Gene
Thanks for pointing this out. Applied in 3b3bc3675519e1530adab27480196eb3f1a458bc Note that I've been adding some code in that branch to support the tftp WRQ from libupload when running with lpxelinux. 2015-09-02 12:18 GMT+02:00 Gene Cumm <gene.cumm at gmail.com>:> On Fri, Aug 14, 2015 at 4:41 AM, Erwan Velu via Syslinux > <syslinux at zytor.com> wrote: > > Hi folks, > > > > I've been off for a pretty long time and I'm trying to resume a little > bit. > > > > While testing HDT I found some serious bugs in HDT's CLI mode but also > took > > the time to update the DMI detection tables. > > > > Please find my hw_update branch which is in sync with the current git > tree. > > https://github.com/ErwanAliasr1/syslinux/commits/hw_update > > f249d92b doesn't feel right. After parsing, it should become an array > of many pointers, one per argument. You do this allocation in > parse_command_line() properly but I believe due to the indirection, > it's not being applied correctly. Don't we need the address of (1) an > array (2) of pointers (3) to blocks of chars? > > > static void parse_command_line(char *line, char **command, char **module, > int *argc, char **argv) > > should probably be > > static void parse_command_line(char *line, char **command, char **module, > int *argc, char ***argv) > > with other appropriate changes while > > parse_command_line(line, &command, &module, &argc, argv); > > should be > > parse_command_line(line, &command, &module, &argc, &argv); > > > Let me know. > > -- > -Gene >