This patch makes PAYLOAD_PREPARSE_ELF only apply to the payload. It also separates the lar steps for adding the payload to make this happen. Because of that, I also separated the nocompress files so that there is no need for that style of flag. I removed it in lar as well.
I rearranged Kconfig a little so that PAYLOAD_PREPARSE_ELF is an option if you've selected to have an ELF payload.
I made the Makefile print PAYLOAD payload.elf -C lzma when the payload is compressed and PAYLOAD payload.elf when it isn't.
Myles
Signed-off-by: Myles Watson mylesgw@gmail.com
On Thu, Feb 21, 2008 at 11:49 AM, Myles Watson mylesgw@gmail.com wrote:
This patch makes PAYLOAD_PREPARSE_ELF only apply to the payload.
oops! I just realized: pre-parsing stage2 is required. So we need to rename this flag and continue to ELF parse stage2.
E.g.:
normal/payload/segment0 (152672 bytes, zeroes compressed to 1 bytes @0x50);loadaddress 0x0x111f60 entry 0x0x10d9b4 normal/payload/segment1 (73544 bytes, lzma compressed to 36370 bytes @0xb0);loadaddress 0x0x100000 entry 0x0x10d9b4 normal/payload/segment2 (72 bytes, lzma compressed to 47 bytes @0x8f20);loadaddress 0x0x1373c0 entry 0x0x10d9b4 normal/option_table (1200 bytes @0x8fa0);loadaddress 0x0 entry 0x0 normal/stage2/segment0 (191788 bytes, zeroes compressed to 1 bytes @0x94a0);loadaddress 0x0xb8e0 entry 0x0x2000 normal/stage2/segment1 (32508 bytes, lzma compressed to 17187 bytes @0x9500);loadaddress 0x0x2000 entry 0x0x2000 normal/stage2/segment2 (6348 bytes, lzma compressed to 425 bytes @0xd880);loadaddress 0x0xa000 entry 0x0x2000 normal/initram/segment0 (5404 bytes @0xda80);loadaddress 0x0 entry 0x0xf0b blob/vsa (57504 bytes @0xefe0);loadaddress 0x0 entry 0x0 zerofill (384767 bytes @0x1d0c0);loadaddress 0x0 entry 0x0 bootblock (20480 bytes @0x7b000) Total size = 523914 bytes (0x7fe8a)
pre-parsing is for more than just payload, it turns out, esp. since it manages bss correctly. It turned out the built-in elf loader did not, which was also a problem as I found out the hard way. E.g. stage2/segment0 above is .bss.
ron
From: ron minnich [mailto:rminnich@gmail.com] On Thu, Feb 21, 2008 at 11:49 AM, Myles Watson mylesgw@gmail.com wrote:
This patch makes PAYLOAD_PREPARSE_ELF only apply to the payload.
oops! I just realized: pre-parsing stage2 is required. So we need to rename this flag and continue to ELF parse stage2.
???
The way the patch does it is to preparse everything except the payload, which depends on the PAYLOAD_PREPARSE_ELF flag. Isn't that what you're saying?
Myles
On 21.02.2008 21:07, Myles Watson wrote:
From: ron minnich [mailto:rminnich@gmail.com] On Thu, Feb 21, 2008 at 11:49 AM, Myles Watson mylesgw@gmail.com wrote:
This patch makes PAYLOAD_PREPARSE_ELF only apply to the payload.
oops! I just realized: pre-parsing stage2 is required. So we need to rename this flag and continue to ELF parse stage2.
???
The way the patch does it is to preparse everything except the payload, which depends on the PAYLOAD_PREPARSE_ELF flag. Isn't that what you're saying?
The patch seems to be OK. Please wait for another ack before commit, though.
Acked-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net
On Thu, Feb 21, 2008 at 12:07 PM, Myles Watson mylesgw@gmail.com wrote:
???
The way the patch does it is to preparse everything except the payload, which depends on the PAYLOAD_PREPARSE_ELF flag. Isn't that what you're saying?
I misread your comment. Sorry.
ron
-----Original Message----- From: ron minnich [mailto:rminnich@gmail.com] Sent: Thursday, February 21, 2008 6:21 PM To: Myles Watson Cc: Coreboot Subject: Re: [coreboot] v3 config patch and lar patch
On Thu, Feb 21, 2008 at 12:07 PM, Myles Watson mylesgw@gmail.com wrote:
???
The way the patch does it is to preparse everything except the payload, which depends on the PAYLOAD_PREPARSE_ELF flag. Isn't that what you're saying?
I misread your comment. Sorry.
ron
No problem.
Myles
The patch seems to be OK. Please wait for another ack before commit, though.
Stefan?
This patch brings us back to where you can choose not to preparse the ELF.
Myles
Acked-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net
On Fri, Feb 22, 2008 at 11:20 AM, Myles Watson mylesgw@gmail.com wrote:
The patch seems to be OK. Please wait for another ack before commit, though.
Stefan?
This patch brings us back to where you can choose not to preparse the ELF.
hmm, I just realized if we have this, we're going to have to repair the coreboot elf parser ... it doesn't do bss correctly.
hmm :-)
ron
-----Original Message----- From: ron minnich [mailto:rminnich@gmail.com] Sent: Friday, February 22, 2008 2:00 PM To: Myles Watson Cc: Carl-Daniel Hailfinger; Coreboot Subject: Re: [coreboot] v3 config patch and lar patch
On Fri, Feb 22, 2008 at 11:20 AM, Myles Watson mylesgw@gmail.com wrote:
The patch seems to be OK. Please wait for another ack before commit, though.
Stefan?
This patch brings us back to where you can choose not to preparse the
ELF.
hmm, I just realized if we have this, we're going to have to repair the coreboot elf parser ... it doesn't do bss correctly.
Maybe I should make it so you have to be in expert mode to not parse the ELF.
Myles
hmm :-)
ron
On 22.02.2008 22:08, Myles Watson wrote:
-----Original Message----- From: ron minnich [mailto:rminnich@gmail.com] Sent: Friday, February 22, 2008 2:00 PM To: Myles Watson Cc: Carl-Daniel Hailfinger; Coreboot Subject: Re: [coreboot] v3 config patch and lar patch
On Fri, Feb 22, 2008 at 11:20 AM, Myles Watson mylesgw@gmail.com wrote:
The patch seems to be OK. Please wait for another ack before commit, though.
Stefan?
This patch brings us back to where you can choose not to preparse the
ELF.
hmm, I just realized if we have this, we're going to have to repair the coreboot elf parser ... it doesn't do bss correctly.
Maybe I should make it so you have to be in expert mode to not parse the ELF.
Add a dependency on BROKEN and stick a FIXME in the coreboot ELF parser. Whoever wants to use the code is free to fix it.
Regards, Carl-Daniel
On Fri, Feb 22, 2008 at 3:21 PM, Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net wrote:
On 22.02.2008 22:08, Myles Watson wrote:
-----Original Message----- From: ron minnich [mailto:rminnich@gmail.com] Sent: Friday, February 22, 2008 2:00 PM To: Myles Watson Cc: Carl-Daniel Hailfinger; Coreboot Subject: Re: [coreboot] v3 config patch and lar patch
On Fri, Feb 22, 2008 at 11:20 AM, Myles Watson mylesgw@gmail.com wrote:
The patch seems to be OK. Please wait for another ack before commit, though.
Stefan?
This patch brings us back to where you can choose not to preparse the
ELF.
hmm, I just realized if we have this, we're going to have to repair the coreboot elf parser ... it doesn't do bss correctly.
Maybe I should make it so you have to be in expert mode to not parse the ELF.
Add a dependency on BROKEN and stick a FIXME in the coreboot ELF parser. Whoever wants to use the code is free to fix it.
How about the middle ground for now and commit this patch, which defaults to parsing the ELF, since that works. Someone who knows where the ELF parser is broken should probably insert the FIXME or fix it.
The only difference between this version and the last is that PARSE_PAYLOAD_ELF defaults to y, and doesn't depend on EXPERT.
Myles
Signed-off-by: Myles Watson mylesgw@gmail.com
Regards, Carl-Daniel
On 22.02.2008 23:58, Myles Watson wrote:
On Fri, Feb 22, 2008 at 3:21 PM, Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net wrote:
On 22.02.2008 22:08, Myles Watson wrote:
-----Original Message----- From: ron minnich [mailto:rminnich@gmail.com] Sent: Friday, February 22, 2008 2:00 PM To: Myles Watson Cc: Carl-Daniel Hailfinger; Coreboot Subject: Re: [coreboot] v3 config patch and lar patch
On Fri, Feb 22, 2008 at 11:20 AM, Myles Watson mylesgw@gmail.com wrote:
The patch seems to be OK. Please wait for another ack before commit, though.
Stefan?
This patch brings us back to where you can choose not to preparse the
ELF.
hmm, I just realized if we have this, we're going to have to repair the coreboot elf parser ... it doesn't do bss correctly.
Maybe I should make it so you have to be in expert mode to not parse the ELF.
Add a dependency on BROKEN and stick a FIXME in the coreboot ELF parser. Whoever wants to use the code is free to fix it.
How about the middle ground for now and commit this patch, which defaults to parsing the ELF, since that works. Someone who knows where the ELF parser is broken should probably insert the FIXME or fix it.
The only difference between this version and the last is that PARSE_PAYLOAD_ELF defaults to y, and doesn't depend on EXPERT.
Myles
Signed-off-by: Myles Watson mylesgw@gmail.com
The coreboot part looks OK and is Acked-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net
I have not reviewed the util/lar part because the code there uses a coding style (not indentation, but structure and functions) I am not familiar with.
Regards, Carl-Daniel
On Fri, Feb 22, 2008 at 4:16 PM, Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net wrote:
On 22.02.2008 23:58, Myles Watson wrote:
On Fri, Feb 22, 2008 at 3:21 PM, Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net wrote:
On 22.02.2008 22:08, Myles Watson wrote:
-----Original Message----- From: ron minnich [mailto:rminnich@gmail.com] Sent: Friday, February 22, 2008 2:00 PM To: Myles Watson Cc: Carl-Daniel Hailfinger; Coreboot Subject: Re: [coreboot] v3 config patch and lar patch
On Fri, Feb 22, 2008 at 11:20 AM, Myles Watson mylesgw@gmail.com wrote:
> The patch seems to be OK. Please wait for another ack before commit, > though.
Stefan?
This patch brings us back to where you can choose not to preparse the
ELF.
hmm, I just realized if we have this, we're going to have to repair the coreboot elf parser ... it doesn't do bss correctly.
Maybe I should make it so you have to be in expert mode to not parse the ELF.
Add a dependency on BROKEN and stick a FIXME in the coreboot ELF parser. Whoever wants to use the code is free to fix it.
How about the middle ground for now and commit this patch, which defaults to parsing the ELF, since that works. Someone who knows where the ELF parser is broken should probably insert the FIXME or fix it.
The only difference between this version and the last is that PARSE_PAYLOAD_ELF defaults to y, and doesn't depend on EXPERT.
Myles
Signed-off-by: Myles Watson mylesgw@gmail.com
The coreboot part looks OK and is Acked-by: Carl-Daniel Hailfinger c-d.hailfinger.devel.2006@gmx.net
Rev 620. Thanks, Myles
I have not reviewed the util/lar part because the code there uses a coding style (not indentation, but structure and functions) I am not familiar with.
Regards, Carl-Daniel
-
Carl-Daniel Hailfinger -
Myles Watson -
ron minnich