support/download: print dl hash if not provided
PATCH 1/1 support/download: print dl hash if not provided
When a hash file exists but the hash is not provided, the script exits without any information about the hash of the downloaded file.
ERROR: No hash found for rpi-userland-771a9aa7155442615bbe4cd6cf87b29b90cd228a.tar.gz
Now, the script hashes the downloaded file and suggests its sha256 to the user.
$ make ... >>> rpi-userland 771a9aa7155442615bbe4cd6cf87b29b90cd228a Downloading --2017-07-19 21:38:39-- https://github.com/raspberrypi/userland/archive/771a9aa7155442615bbe4cd6cf87b29b90cd228a/rpi-userland-771a9aa7155442615bbe4cd6cf87b29b90cd228a.tar.gz ... ERROR: No hash found for rpi-userland-771a9aa7155442615bbe4cd6cf87b29b90cd228a.tar.gz ERROR: If the source is trusted, consider adding these lines to package/rpi-userland//rpi-userland.hash # Locally calculated from download sha256 771fb1be53414b00a9213f24bd9c88059cf76b72c5e21ac613b267d3e58d3715 rpi-userland-771a9aa7155442615bbe4cd6cf87b29b90cd228a.tar.gz ...
It also fixes check_one_hash description. check_one_hash() takes three arguments:
- algo hash
- known hash
- file to hash
--- support/download/check-hash | 29 +++++++++++++++++++++++------ 1 file changed, 23 insertions(+), 6 deletions(-)
support/download/check-hash +23 -6
@@ -41,13 +41,12 @@ if [ ! -f "${h_file}" ]; then exit 0 fi -# Check one hash for a file -# $1: known hash +# Compute hash for a file +# $1: algo hash # $2: file (full path) -check_one_hash() { +compute_hash() { _h="${1}" - _known="${2}" - _file="${3}" + _file="${2}" # Note: md5 is supported, but undocumented on purpose. # Note: sha3 is not supported, since there is currently no implementation
@@ -66,8 +65,20 @@ check_one_hash() { ;; esac + ${_h}sum "${_file}" |cut -d ' ' -f 1 +} + +# Check one hash for a file +# $1: algo hash +# $2: known hash +# $3: file (full path) +check_one_hash() { + _h="${1}" + _known="${2}" + _file="${3}" + # Do the hashes match? - _hash=$( ${_h}sum "${_file}" |cut -d ' ' -f 1 ) + _hash=$( compute_hash "${_h}" "${_file}" ) if [ "${_hash}" = "${_known}" ]; then printf "%s: OK (%s: %s)\n" "${base}" "${_h}" "${_hash}" return 0
@@ -105,6 +116,12 @@ if [ ${nb_checks} -eq 0 ]; then exit 0 ;; esac + + h="sha256" + hash=$( compute_hash "${h}" "${file}" ) printf "ERROR: No hash found for %s\n" "${base}" >&2 + printf "ERROR: If the source is trusted, consider adding these lines to ${h_file}\n" >&2 + printf "# Locally calculated from download\n" >&2 + printf "${h} ${hash} ${base}\n" >&2 exit 3 fi
Signature · 1 line
2.13.2
PATCH 1/1 support/download: print dl hash if not provided
Ga?l, All,
On 2017-07-19 23:18 -0400, Ga?l PORTAY spake thusly:
Quoted 19 lines
When a hash file exists but the hash is not provided, the script exits without any information about the hash of the downloaded file.
ERROR: No hash found for rpi-userland-771a9aa7155442615bbe4cd6cf87b29b90cd228a.tar.gzNow, the script hashes the downloaded file and suggests its sha256 to the user.
$ make ... >>> rpi-userland 771a9aa7155442615bbe4cd6cf87b29b90cd228a Downloading --2017-07-19 21:38:39-- https://github.com/raspberrypi/userland/archive/771a9aa7155442615bbe4cd6cf87b29b90cd228a/rpi-userland-771a9aa7155442615bbe4cd6cf87b29b90cd228a.tar.gz ... ERROR: No hash found for rpi-userland-771a9aa7155442615bbe4cd6cf87b29b90cd228a.tar.gz ERROR: If the source is trusted, consider adding these lines to package/rpi-userland//rpi-userland.hash # Locally calculated from download sha256 771fb1be53414b00a9213f24bd9c88059cf76b72c5e21ac613b267d3e58d3715 rpi-userland-771a9aa7155442615bbe4cd6cf87b29b90cd228a.tar.gz ...
Quoted 7 lines
It also fixes
check_one_hashdescription.check_one_hash()takes three arguments:
- algo hash
- known hash
- file to hashSigned-off-by Ga?l PORTAY <[email protected]>
NAK from me.
The reason we do not want this is that we instead want the user to go fetch the hash(es) as provided by upstream, like in an announcement email, or in an on-the-side hash file.
Having the download infra print the locally computed hash defeats the very purpose of hashes: check that we get what upstream provides.
We only accept local calculations of hashes for the cases where upstream does not provide any (or too weak) hash.
As an aside, this patch does two things: fix the comment for check_one_hash() and print the hash. It should be split.
Regards,
Yann E. MORIN.
Quoted 64 lines
--- support/download/check-hash | 29 +++++++++++++++++++++++------ 1 file changed, 23 insertions(+), 6 deletions(-)support/download/check-hash +23
-6@@ -41,13 +41,12 @@ if [ ! -f "${h_file}" ]; then exit 0 fi -# Check one hash for a file -# $1: known hash +# Compute hash for a file +# $1: algo hash # $2: file (full path) -check_one_hash() { +compute_hash() { _h="${1}" - _known="${2}" - _file="${3}" + _file="${2}" # Note: md5 is supported, but undocumented on purpose. # Note: sha3 is not supported, since there is currently no implementation@@ -66,8 +65,20 @@ check_one_hash() { ;; esac + ${_h}sum "${_file}" |cut -d ' ' -f 1 +} + +# Check one hash for a file +# $1: algo hash +# $2: known hash +# $3: file (full path) +check_one_hash() { + _h="${1}" + _known="${2}" + _file="${3}" + # Do the hashes match? - _hash=$( ${_h}sum "${_file}" |cut -d ' ' -f 1 ) + _hash=$( compute_hash "${_h}" "${_file}" ) if [ "${_hash}" = "${_known}" ]; then printf "%s: OK (%s: %s)\n" "${base}" "${_h}" "${_hash}" return 0@@ -105,6 +116,12 @@ if [ ${nb_checks} -eq 0 ]; then exit 0 ;; esac + + h="sha256" + hash=$( compute_hash "${h}" "${file}" ) printf "ERROR: No hash found for %s\n" "${base}" >&2 + printf "ERROR: If the source is trusted, consider adding these lines to ${h_file}\n" >&2 + printf "# Locally calculated from download\n" >&2 + printf "${h} ${hash} ${base}\n" >&2 exit 3 fiSignature · 1 line
2.13.2
Signature · 6 lines
.-----------------.--------------------.------------------.--------------------. | Yann E. MORIN | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: | | +33 662 376 056 | Software Designer | \ / CAMPAIGN | ___ | | +33 223 225 172 `------------.-------: X AGAINST | \e/ There is no | | http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL | v conspiracy. | '------------------------------^-------^------------------^--------------------'
PATCH 1/1 support/download: print dl hash if not provided
Yann,
Quoted 19 lines · Yann E. MORIN
Ga?l, All,
On 2017-07-19 23:18 -0400, Ga?l PORTAY spake thusly:
Quoted 9 lines
...
It also fixes
check_one_hashdescription.check_one_hash()takes three arguments:
- algo hash
- known hash
- file to hashSigned-off-by Ga?l PORTAY <[email protected]>NAK from me.
The reason we do not want this is that we instead want the user to go fetch the hash(es) as provided by upstream, like in an announcement email, or in an on-the-side hash file.
Quoted 6 lines
Having the download infra print the locally computed hash defeats the very purpose of hashes: check that we get what upstream provides.
We only accept local calculations of hashes for the cases where upstream does not provide any (or too weak) hash.
Okay.
Quoted 3 lines
As an aside, this patch does two things: fix the comment for
check_one_hash()and print the hash. It should be split.
I will send a patch for this tiny nitpick.
Regards,
Ga?l
PATCH 1/1 support/download: print dl hash if not provided
Hi Ga?l,
Quoted 22 lines · Ga?l PORTAY
Yann,
Quoted 19 lines · Yann E. MORIN
Ga?l, All,
On 2017-07-19 23:18 -0400, Ga?l PORTAY spake thusly:
Quoted 9 lines
...
It also fixes
check_one_hashdescription.check_one_hash()takes three arguments:
- algo hash
- known hash
- file to hashSigned-off-by Ga?l PORTAY <[email protected]>NAK from me.
The reason we do not want this is that we instead want the user to go fetch the hash(es) as provided by upstream, like in an announcement email, or in an on-the-side hash file.
Quoted 8 lines
Quoted 6 lines
Having the download infra print the locally computed hash defeats the very purpose of hashes: check that we get what upstream provides.
We only accept local calculations of hashes for the cases where upstream does not provide any (or too weak) hash.
Okay.
Thomas and I discussed this at the BR developer meeting, and we disagree with
Yann that we should make life difficult for people bumping a package :-P. So we think this patch does have value. Would you be willing to respin it?
However, the text you propose is not strong enough. How about:
Please find a hash in the upstream announcement or website and add it to ${h_file}
If upstream doesn't provide a hash and the source is trusted, consider adding these lines:
Also, the most annoying thing actually is that when the hash is wrong, the
just-downloaded file will be removed again. It would be convenient to avoid removing it, similar to how it is done when the file exists already.
Regards,
Arnout
Quoted 14 lines
Quoted 4 lines
As an aside, this patch does two things: fix the comment for
check_one_hash()and print the hash. It should be split.I will send a patch for this tiny nitpick.
Regards,
Ga?l
_______________________________________________
buildroot mailing list
buildroot at busybox.net
http://lists.busybox.net/mailman/listinfo/buildroot
Signature · 6 lines
Arnout Vandecappelle arnout at mind be Senior Embedded Software Architect +32-16-286500 Essensium/Mind http://www.mind.be G.Geenslaan 9, 3001 Leuven, Belgium BE 872 984 063 RPR Leuven LinkedIn profile: http://www.linkedin.com/in/arnoutvandecappelle GPG fingerprint: 7493 020B C7E3 8618 8DEC 222C 82EB F404 F9AC 0DDF
PATCH 1/1 support/download: print dl hash if not provided
Hi Arnout,
Quoted 32 lines · Arnout Vandecappelle
Hi Ga?l,
Quoted 29 lines · Ga?l PORTAY
Yann,
Quoted 25 lines · Yann E. MORIN
Ga?l, All,
On 2017-07-19 23:18 -0400, Ga?l PORTAY spake thusly:
Quoted 9 lines
...
It also fixes
check_one_hashdescription.check_one_hash()takes three arguments:
- algo hash
- known hash
- file to hashSigned-off-by Ga?l PORTAY <[email protected]>NAK from me.
The reason we do not want this is that we instead want the user to go fetch the hash(es) as provided by upstream, like in an announcement email, or in an on-the-side hash file.
Having the download infra print the locally computed hash defeats the very purpose of hashes: check that we get what upstream provides.
We only accept local calculations of hashes for the cases where upstream does not provide any (or too weak) hash.
Quoted 6 lines
Quoted 1 line
Okay.
Thomas and I discussed this at the BR developer meeting, and we disagree with
Yann that we should make life difficult for people bumping a package :-P. So we think this patch does have value. Would you be willing to respin it?
For sure. Let me do a rebase and I will respin it at the end of the day.
Quoted 7 lines
However, the text you propose is not strong enough. How about:
Please find a hash in the upstream announcement or website and add it to ${h_file}
If upstream doesn't provide a hash and the source is trusted, consider adding these lines:
I agree.
Quoted 4 lines
Also, the most annoying thing actually is that when the hash is wrong, the
just-downloaded file will be removed again. It would be convenient to avoid removing it, similar to how it is done when the file exists already.
I was thinking the same thing when write this patch. Let me see. I think it is just another case to add in the download script.
Regards,
Gael